2
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
About KAPSARC
The King Abdullah Petroleum Studies and Research Center (KAPSARC) is a
non-prot global institution dedicated to independent research into energy economics,
policy, technology and the environment across all types of energy. KAPSARC’s
mandate is to advance the understanding of energy challenges and opportunities
facing the world today and tomorrow, through unbiased, independent, and high-caliber
research for the benet of society. KAPSARC is located in Riyadh, Saudi Arabia.
This publication is also available in Arabic.
Legal Notice
© Copyright 2021 King Abdullah Petroleum Studies and Research Center (“KAPSARC”).
This Document (and any information, data or materials contained therein) (the
“Document”) shall not be used without the proper attribution to KAPSARC. The
Document shall not be reproduced, in whole or in part, without the written permission
of KAPSARC. KAPSARC makes no warranty, representation or undertaking whether
expressed or implied, nor does it assume any legal liability, whether direct or indirect,
or responsibility for the accuracy, completeness, or usefulness of any information that
is contained in the Document. Nothing in the Document constitutes or shall be implied to
constitute advice, recommendation or option. The views and opinions expressed in this
publication are those of the authors and do not necessarily reect the ofcial views or
position of KAPSARC.
3
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
Key Points
Electricity demand in Saudi Arabia is undergoing unprecedented changes following the
implementation of efciency measures and energy price reforms. For the rst time in decades,
the country’s electricity demand is stagnating, suggesting that consumer behavior has structurally
shifted. These changes raise uncertainties about the potential trajectory of long-term electricity demand.
Thus, this study uses a computable general equilibrium model to project sectoral electricity demand
in Saudi Arabia through 2030. We project that growth in total Saudi electricity demand will signicantly
decelerate over the coming decade compared with historical trends. In our reference scenario, this demand
reaches 365.4 terawatthours (TWh) by 2030. However, our sectoral decomposition shows large disparities
across sectors. Demand is projected to grow more rapidly in the industrial and services segments than in
the residential sector. Nevertheless, the latter will still account for the largest share of total consumption
in 2030. We also simulate four additional scenarios for domestic electricity price reforms and efciency
policies. Successfully implementing these measures may result in signicant energy savings. Aligning
Saudi electricity prices with the average electricity price among G20 countries can reduce total electricity
demand by up to 71.6 TWh in 2030. Independently enforcing efciency policies can reduce total electricity
demand by up to 118.7 TWh. Moreover, alternative policy scenarios suggest that the macroeconomic gains
from energy savings can alleviate some of the Saudi energy system’s burden on public nance.
We use a computable general equilibrium model to project Saudi electricity demand through 2030.
Saudi electricity demand is expected to grow more slowly over the coming decade relative to its
historical trend.
The timing and size of recently announced large-scale projects may increase our demand projections,
and these effects may be analyzed in future research.
We simulate various hypothetical scenarios for price reforms and efciency gains in electricity
intensity uses.
Price reforms and efciency measures may reduce total demand by 11% to 32% in 2030, with higher
savings realized under energy efciency measures.
The macroeconomic gains from energy efciency scenarios are greater than those from price
reform scenarios.
4
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
1. Introduction
Projecting future demand for electricity is
central to power sector planning, as these
projections inform capacity investment
requirements and related infrastructure expansions.
Electricity is not currently economically storable
in large volumes. Thus, the underlying drivers of
electricity demand and potential market shifts must
be carefully considered to minimize power system
costs.1
In the Kingdom of Saudi Arabia, demand for
electricity has grown rapidly since the development
of the electricity sector in the early 1970s. This
growth has been driven by a rapidly increasing
population, dynamic economic growth and low
regulated energy prices. In 2018, total Saudi
electricity demand reached 299.2 terawatthours
(TWh).2 Saudi Arabia is the 14-largest electricity
consumer in the world. Its consumption is similar
to that of more populated countries (e.g., Mexico,
whose 2019 population was 127.5 million, compared
to 34.2 million for Saudi Arabia). It is also on par
with more advanced economies (e.g., Italy, whose
2019 gross domestic product (GDP) was $2,151.4
billion, compared to $704.0 billion for Saudi Arabia),
according to The World Bank.
In recent years, the Saudi government has
addressed the rapidly increasing fuel consumption
of its power sector by expanding the use of efcient
gas plants. This step has reduced the country’s
reliance on oil and rened products for power
generation. Moreover, Saudi policymakers have also
enacted some demand-side measures. In 2010,
the Kingdom began promoting several efciency
initiatives to rationalize energy consumption with the
establishment of the Saudi Energy Efciency Center
(SEEC 2018). Additionally, the Saudi government
implemented the rst round of national energy price
reforms (EPR) in 2016, with a second round in 2018.
The scale of these recently implemented EPR and
efciency measures is unprecedented in Saudi
Arabia. Thus, these policies’ potential effects
on future demand cannot be assessed based
on past experiences. Instead, it is necessary to
strengthen the methodological aspects of energy
demand projections. Using advanced analytical
tools to capture market transformations, behavioral
adjustments and interdependencies across
economic agents, we can better project electricity
demand pathways. In doing so, we build upon the
work of Soummane et al. (2022). They develop
a hybrid energy-economy computable general
equilibrium (CGE) model that accounts for specic
features of the Saudi economy. These features
include administered domestic energy prices and a
currency peg to the United States (U.S.) dollar.
The remainder of this paper is organized as follows.
In section 2, we review the literature related to Saudi
electricity demand and applications of general
equilibrium models in energy-related studies. In
section 3, we describe our projection methodology
and present the sectoral components of Saudi
electricity demand. Section 4 summarizes the
scenarios and the underlying assumptions for the
demand projections. Section 5 presents the main
results of our analysis through 2030, and section 6
concludes the paper.
5
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
2. Literature review
2.1. Studies on Saudi
electricity demand
To the best of our knowledge, Hasanov (2019)
is the only published study that provides partial
projections of Saudi electricity demand. Unlike our
study, Hasanov (2019) focuses on industrial demand
and uses a time horizon of 2025. The author models
annual industrial electricity demand from 1984 to
2016. He uses the real price of electricity, industrial
value-added, the working-age population and the
cost of capital as explanatory variables. The model
shows that industrial electricity demand is relatively
price and income inelastic, with long-run elasticities
of -0.1 and within 0.2-0.3, respectively. Furthermore,
the estimations show that active population
dynamics and the cost of capital are signicant
drivers of industrial electricity demand.
Additionally, several previous econometric studies
aim to identify the drivers of Saudi electricity
demand, focusing on its responsiveness to prices
and income. Al-Sahlawi (1990) and Diabi (1998) both
estimate aggregate Saudi demand for electricity as
a function of income (proxied by real GDP) and real
electricity prices. Both studies conclude that Saudi
electricity demand is income and price inelastic.
Likewise, Al-Faris (2002) models aggregate Saudi
electricity demand from 1970 to 19973 as a function
of income (proxied by real GDP) and real electricity
prices. He also includes the price of liqueed
petroleum gas, an alternative fuel, to capture
substitution effects. He nds that Saudi electricity
demand is insensitive to price changes but elastic to
income, with income elasticities of 0.05 in the short
run and 1.65 in the long run.
Atalla and Hunt (2016) model residential electricity
demand in GCC countries using annual data from
1985 to 2012. Like other studies, they include the
real price of electricity and real GDP as a proxy
for income in their model. They also include the
population and weather conditions, captured by
heating and cooling degree-days. They nd that
electricity demand is relatively inelastic to income
and prices in the short and long terms. The absolute
values of the corresponding elasticities are less than
0.5. However, the population and weather conditions
have more signicant short- and long-run impacts.
Lastly, two post-EPR studies model Saudi residential
electricity demand to estimate price and income
elasticities. Aldubyan and Gasim’s (2020) model
includes electricity prices, real GDP per capita as an
income proxy and cooling degree-days to capture
weather effects. They conrm that residential
electricity demand is price and income inelastic,
as they estimate long-term elasticities of -0.09 and
0.22, respectively. However, their decomposition
analysis shows that the electricity price hike in 2018
was the key contributor to the observed drop in the
demand that year.
Mikayilov et al. (2020) similarly use prices, income
and weather to explain residential electricity demand
across four regions of Saudi Arabia. Their results
conrm that aggregate residential demand is
inelastic to price and income changes. However,
the results vary signicantly across the regions.
For instance, the long-term price elasticities range
from -0.2 in the central region to -0.5 in the eastern
region. Likewise, the long-term income elasticities
range from 0.3 in the eastern region to 1.0 in the
western region.
We can draw two conclusions from these studies.
First, most prior studies develop econometric
estimations of Saudi electricity demand. Moreover,
they generally focus on aggregate demand or on
one segment of aggregate demand (e.g., residential
or industrial demand). To the best of our knowledge,
no prior study has modeled or projected Saudi
electricity demand for several sectors
6
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
2. Literature review
simultaneously.4 Second, few studies investigate
the post-EPR period. However, as described above,
the EPR initiated fundamental shifts in electricity
demand patterns, and these shifts are likely to
persist in the future. These changes must be
considered when modeling future demand.
Thus, this study lls two gaps in the existing
literature on Saudi electricity demand. First,
it models and projects sectoral demand in a
consistent framework, capturing interdependencies
across sectors through the general equilibrium
approach. Second, it accounts for recent changes in
electricity demand patterns driven by price reforms
and efciency measures.
2.2. Energy and electricity
modeling using the CGE
framework
Since Johansen’s (1960) seminal work, general
equilibrium modeling has been widely used to
assess economic, energy and environmental
trajectories in the context of signicant changes.
The purpose of a CGE model is to provide a
comprehensive estimation of a policy’s effects.
Incorporating consumption and production
functions within a multi-sector and multi-market
framework allows for better estimations of supply
and demand relative to partial equilibrium models.
CGE models can capture market interactions,
penetrations of price shifts in the input-output
matrix and consecutive adjustments of input
trade-offs. Ultimately, they can capture household
consumption.
Most recent studies applying CGE frameworks in
an energy context estimate the energy demand
response to price shifts, production mix evolutions
and subsidy reforms. Lin and Jiang (2011), Liu
and Li (2011) and Chi et al. (2014) investigate the
impacts of subsidy reforms on energy demand and
macroeconomic indicators in China. Kat et al. (2018)
use a CGE framework to analyze the prospects
of various energy scenarios in Turkey and the
underlying emissions trajectories. Böhringer and
Rutherford (2013) explore energy and emission
scenarios for Poland. One recent application of a
CGE model to Kuwait, a GCC country, shows that
energy subsidy reforms have benecial effects on
economic diversication (Shehabi 2020). Another
such study nds that using direct transfers to
compensate for electricity subsidy reforms can
offset welfare losses (Gelan 2018).
Holmøy (2005) and He et al. (2011) apply
CGE models to investigate electricity demand.
Specically, they estimate the sensitivity of
electricity demand to changes in electricity prices
in Norway and China, respectively. These studies
emphasize the prominent role of substitution
elasticities of demand functions. Beckman et al.
(2011) show that an adequately parameterized CGE
model replicates historic energy demand and supply
trajectories well.
Several studies use the CGE framework to
investigate specic policy issues related to the
Kingdom. These issues include behavior within the
oil market (De Santis 2003), the abrogation of trade
tariffs (Al-Hawwas 2010) and exchange rate policies
(Al-Thumairi 2012). Soummane et al. (2019) and
Soummane et al. (2022) recently developed a CGE
model for the Kingdom that acknowledges certain
features of the Saudi economy. This model includes
administered energy prices within a detailed
representation of the energy sector to assess
economic diversication outcomes.
7
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
3. Methodology
3.1. CGE model with
improved electricity
demand characteristics
This study employs IMACLIM-SAU, an
economy-wide dynamic CGE model that
embodies specic features of the Saudi
economy. We extend Soummane et al. (2019)
and Soummane et al. (2022) by improving the
model’s representation of electricity demand.
The model covers 13 sectors, as reported in
Table 1. To be concise, we describe only the
model features related to the electricity sector
in this section. In Annex A, we summarize the
main socioeconomic drivers of the scenarios
that we explore. Soummane et al. (2022) provide
a detailed algebraic description, calibration
procedures and data treatments for the model.
Table 1. Sectoral coverage of IMACLIM-SAU.
Abbreviation Sector
Energy
OIL Crude oil
GAS Natural gas
RFN Rening
ELE Electricity
Non-energy
AGR Agriculture, hunting, forestry and shing
MIN Other mining (excluding oil and gas extraction)
CHM Chemicals and petrochemicals
NMM Non-metallic minerals (including cement)
MAN Manufacturing
PRV Private sector services
PUB Public services
TRA Transport: air and sea
OTP Other transport
8
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
3. Methodology
Our model maintains three important specications
of Soummane et al. (2019) and Soummane et
al. (2022). We maintain the model’s closure and
relationship between the trade balance and
real effective exchange rate. We also maintain
the correlation between the average wage and
unemployment rate.
However, our model departs from that of
Soummane et al. (2022) in two ways to better
represent electricity demand. First, we adjust
the sectoral denitions to separate private and
government services. The latter consume a
signicant amount of electricity, comprising
13.2% of total demand on average between 2013
and 2018. Second, we assume that households’
consumption of energy goods (i.e., rened products
and electricity) is income and price elastic.
The original model treats this consumption as
exogenous (imported from bottom-up expertise).
For non-energy goods, we maintain Soummane et
al.’s (2022) formulation of constant shares of the
budget remainder (i.e., the budget net of energy
expenses).
The consumption function that we adopt for
residential electricity demand helps address
the shortcomings of using GDP as a proxy for
income. Indeed, Atalla and Hunt (2016) highlight
that household income is more appropriate than
GDP per capita in this setting. In GCC countries,
GDP per capita is highly correlated with oil prices.5
Similar to Le Treut (2017), we formulate household
consumption of energy goods Ci as follows:6
(1)
Here, σCPi and σCRi are the price and income
elasticities, respectively. pCi and Rc are the consumer
price of energy good i and consumed income,
respectively. An index of 0 denotes the calibration
value of a variable, that is, the 2013 value.
For households, oil and gas consumption are
essentially nil, and thus, we only need the values of
price and income elasticities for ELE and RFN. We
derive these values from the estimates of Hasanov
et al. (2020) and Mikayilov et al. (2019), respectively.
Thus, for ELE, the income and price elasticities are
set to 0.33 and -0.13, respectively, and for RFN, they
are set to 0.13 and -0.27, respectively. We compute
household incomes using assumptions regarding
the macroeconomic income distribution that are
described in Annex A.2 of Soummane et al. (2022).
The production function of goods and services,
including electricity,7 takes a nested form. To
simulate distinctive scenarios for price reforms
and intensity gains, the model includes two
alternative production specications. In the rst
(Specication 1), capital, labor and electricity
are substitutable inputs in the lower stage of
the production function. They are incorporated
in a constant elasticity of substitution function
to create electried value-added.8 In the upper
stage, electried value-added is combined with
all energy products except ELE to produce a
composite good ( VA_ E) based on a Leontief
function. VA_ E is then combined with materials
(i.e., non-energy products, denoted as M) to
produce domestic output Y. In this specication,
electricity use intensities (i.e., units of ELE per
unit of Y) are endogenously determined from the
modeled prices of electricity and other factors.
This specication is common to all sectors.
The second specication (Specication 2) uses
a similar nested production function. However, in
this specication, electricity intensities, like the
intensities of other energy goods, are determined
exogenously. The intensities of other energy goods
remain constant at their calibration year levels.
In both specications, regulated energy prices
(including ELE prices) are implemented using
agent-specic margins to reect differences
∀∈{,,,}! =3#!"
$%& '
#!"#4(!$"5)%
$%& '
)%#6(!&"!#.
(1)
∀∈{,,,}! =3#!"
$%& '
#!"#4(!$"5)%
$%& '
)%#6(!&"!#. (1)
9
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
3. Methodology
in agent-specic tariffs using output costs. This
implementation is presented in Annex B.6 of
Soummane et al. (2022).
3.2. Simulation of 2014-2018
Saudi electricity demand
One way to validate CGE models and increase
their credibility is to test their performance
against historical data (Devarajan and Robinson
2002). Thus, in this section, we test our model’s
ability to replicate sectoral electricity demand for
2014-2018. As Soummane et al. (2022) describe,
IMACLIM-SAU is calibrated on the base year of
2013. Additionally, the period from 2014 to 2017
is used to calibrate the macroeconomic variables
(i.e., GDP, capital dynamics, the unemployment
rate and the trade balance). In this study, we
maintain this specication.
For residential electricity demand, we implement
the price- and income-elastic demand
function given by Equation 1. For intermediate
electricity uses (i.e., industrial, commercial and
government electricity demand), we implement
the two specications described in the previous
subsection. In the rst specication, electricity
intensities are adapted to regulated prices to
reect producers’ input trade-offs. In the second
specication, we set electricity intensities
exogenously to reect 2014-2018 electricity
consumption per unit of sectoral output, following
SAMA (2019). Thus, the test for this specication
indicates our model’s ability to replicate sectoral
activity levels.9 For agriculture, we keep the
electricity intensity constant over the calibration
years in both specications given its small
share of electricity consumption (2% in 2013).
Electricity prices are average sectoral prices
based on observed regulated tariffs weighted
by consumption brackets, as Soummane (2021)
presents.
For both specications, the simulated total demand
is, on average, within 2.8% of the observed total
demand. Residential demand, which is simulated
using the consumption function with price and
income elasticities, captures the effects of the
slower income increase and EPR. The estimated
trajectories again uctuate around the observed
values by 2.8% on average. For industrial,
commercial, governmental and agricultural use,
the modeled consumption values replicate the
actual values fairly closely. The deviations from
the actual values are below 11% on average in all
segments under both specications. The slight
overestimations of industrial and commercial
demand are compensated by the underestimation of
governmental demand (Figure 1).
10
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
3. Methodology
Figure 1. Modeled versus observed electricity demand across sectors, 2013-2018.
Sources: ECRA (2018) for observed demand and IMACLIM-SAU results for modeled trajectories. Specication 1 uses endogenous
electricity intensities, and specication 2 uses exogenous electricity intensities. Estimated versus observed sectoral activities are the
remaining sources of discrepancy in specication 2.
0
20
40
60
80
100
120
140
160
2013 2014 2015 2016 2017 2018
TWh
Residential
0
10
20
30
40
50
60
70
2013 2014 2015 2016 2017 2018
TWh
Industrial
150
170
190
210
230
250
270
290
310
330
2013 2014 2015 2016 2017 2018
TWh
Total demand
0
10
20
30
40
50
2013 2014 2015 2016 2017 2018
TWh
Government
0
1
2
3
4
5
6
7
2013 2014 2015 2016 2017 2018
TWh
Agriculture
0
10
20
30
40
50
60
70
80
2013 2014 2015 2016 2017 2018
TWh
Commercial
Observed Specification 2Observed Specification 1 Specification 2
Observed Specification 1 Specification 2Observed Specification 1 Specification 2
Observed Specification 1 Specification 2Observed Specification 1 Specification 2
11
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
4. Electricity demand scenarios
through 2030
This section describes our projection scenarios
and their underlying drivers and assumptions.
For sectoral demand projections, we
consider three scenarios reecting Saudi electricity
demand pathways under price reforms and energy
efciency measures to reduce electricity intensities.
We simulate a reference scenario (REF) and two
alternative scenarios. These alternative scenarios
are called the price reform scenario (PR) and energy
efciency scenario (EE). The three scenarios use
similar trajectories for socioeconomic indicators (i.e.,
the active population, labor productivity, oil prices,
oil output, investment and non-energy exports).
They differ only in aspects related to the electricity
sector, as the following subsections explain. Annex
A describes the socioeconomic indicators in more
detail.
4.1. Reference scenario
In this scenario, intermediate and nal electricity
prices remain at their 2018 levels through 2030.
For the residential sector, prices are 139% greater
than pre-EPR levels. For industry (IND), which
corresponds to MIN, CHM, NMM and MAN in our
nomenclature (Table 1), prices are 20% greater than
pre-EPR levels. For PRV, PUB and AGR, prices are
20%, 30% and 50% greater than pre-EPR levels,
respectively. The projected REF scenario is based
on Specication 2; that is, the specication with
exogenous electricity intensities. This specication
deviates less from the observed 2013-2018 demand
than Specication 1 does (see section 3.2).10 In this
scenario, we assume that the enforced intensity
levels for IND, PRV, PUB and AGR remain constant
at their 2013-2018 averages through 2030. This
assumption avoids unduly placing a higher weight
on the intensity in any specic year.
4.2. Price reform scenarios
In this scenario, we consider additional price
reforms beyond the two EPR rounds that have
already taken place. The Saudi government’s Fiscal
Balance Program plans to progressively increase
energy prices to meet “market levels.” This plan has
been highlighted as Saudi Arabia’s most important
initiative (Kingdom of Saudi Arabia 2017; Kingdom of
Saudi Arabia 2019).
This scenario builds on Specication 1 of the
production function (see section 3.1). This
specication endogenizes electricity intensities to
allow for trade-offs with value-added if electricity
prices increase. We investigate two variants of
price increases based on different international
references. In the rst variant, denoted as PR-EM,
electricity prices converge to the average price
across emerging countries in the G20, a group of
leading rich and developing nations. In the second
variant, denoted as PR-AVG, electricity prices
converge to the average price across all G20
countries.11
In the rst variant (PR-EM), residential prices
grow 72% higher than their post-EPR levels. They
ultimately approach current prices in China and
South Africa. Nevertheless, they remain two times
lower than prices in countries with similar incomes
(i.e., GDP per capita) to Saudi Arabia. Examples
of such countries are the Czech Republic, Poland
and Slovakia. Reforming the industrial segment’s
energy prices is a sensitive issue. Saudi Arabia has
a large (mainly energy-intensive) industrial base that
is integrated with the crude oil and natural gas value
chain. However, additional moderate reforms are
necessary to support the economic viability of the
country’s power system (Anouti et al. 2020).
12
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
4. Electricity demand scenarios through 2030
We seek to reect this tension between
reforming electricity prices and preserving local
industrial competitiveness. Thus, we exclude
the two countries with the highest industrial
electricity prices, Brazil and India, from the
computation of the prevailing industrial price in
the PR-EM variant. As a result, the industrial
electricity price in 2030 is 43% higher than the
post-EPR price, which is maintained in the REF
scenario. Hasanov (2019) applies a similar price
increase to the Saudi industrial segment based
on matching the U.S. industrial electricity price
for 2008-2017. For the remaining sectors, that
is, PRV, PUB and AGR, we assume a similar
conservative electricity price increase of 43%
from post-EPR levels by 2030.
The second variant (PR-AVG) assumes that
residential prices triple from their 2018 levels by
2030. As in the PR-EM variant, we exclude the
countries with the two highest electricity prices
(now Italy and Japan) from the computation
for industrial prices. Under this assumption,
industrial prices double by 2030 from their
current levels. This price increase is again
similar to the targeted industrial price in the
Hasanov (2019) alternative scenario.12 Finally,
we assume that the other sectoral prices double
by 2030 compared to their 2018 levels. Table 2
summarizes our assumptions for the PR.
Table 2. Assumptions for the two price reform scenarios (Saudi riyals [SAR]/kilowatthour).
Sources: The authors’ computations of 2013 and 2018 data are based on Nachet and Aoun (2015), Kingdom of Saudi Arabia (2017),
APICORP (2018) and Hasanov (2019). The assumptions for projected values are based on data from Enerdata (2021).
Variant Sector 2013 2018 2025 2030 ∆ 2018-2030
PR-EM
Residential 0.08 0.19 0.25 0.32 +72%
Industrial 0.15 0.18 0.22 0.26 +43%
Commercial 0.22 0.26 0.32 0.37 +43%
Government 0.26 0.32 0.39 0.46 +43%
Agriculture 0.11 0.17 0.21 0.24 +43%
PR-AVG
Residential 0.08 0.19 0.35 0.56 +200%
Industrial 0.15 0.18 0.27 0.36 +100%
Commercial 0.22 0.26 0.39 0.52 +100%
Government 0.26 0.32 0.48 0.64 +100%
Agriculture 0.11 0.17 0.25 0.34 +100%
13
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
4. Electricity demand scenarios through 2030
Finally, we highlight one important feature of our
model of the electricity sector in this scenario. In
our model, electricity is treated as one homogenous
good. In other words, its production is a top-down
function, meaning that we do not differentiate
between discrete generation technologies from a
bottom-up perspective. This approach provides
valuable indications of consumers’ reactions to
price changes. However, it cannot comprehensively
illustrate the evolution of the supply-side mix or the
costs facing technical constraints. Wing (2006),
Cai and Arora (2015) and Kat et al. (2018) describe
ways to integrate electricity sector technology
within a CGE framework in detail. We consider this
integration as a potential direction for future work on
the Saudi electricity market.
4.3. Energy efciency
scenarios
In this scenario, changes in sectoral demand relative
to the REF scenario are driven by decreases in
the electricity intensities of different production
types. Saudi authorities have set energy efciency
measures targeting various power-consuming
segments to contain electricity demand growth.13
The government established the SEEC to set
and coordinate national programs to rationalize
energy consumption in buildings, industry and
transportation. These sectors are responsible for
90% of domestic energy consumption (SEEC 2018).
Furthermore, the national utility company, the Saudi
Electricity Company (SEC), has taken various
measures to rationalize power consumption in the
industry and services sectors. These measures are
an application of the national strategy to reform the
energy sector (Kingdom of Saudi Arabia 2017).14
The energy efciency (EE) scenario uses
Specication 2 for the production function (see
section 3.1), that is, we assume that the intensities
of electricity use are exogenous. We simulate two
variants of the EE: Moderate (EE-Mod) and High
(EE-High). In these variants, we adjust the sectors’
exogenous electricity intensities (i.e., the electricity
demand per unit of output). This specication
therefore comprises the industry (RFN, MIN, CHM,
NMM and MAN), commercial (PRV), government
(PUB) and agriculture (AGR) sectors. We set
residential electricity consumption based on the
assumptions described below.
In the EE-Mod variant, we assume that the
electricity intensity for IND decreases by 14%
relative to 2018 by 2030. In other words, it
decreases by 1.2% per year on average over
this period. This decline corresponds to half of
the electricity intensity improvement achieved
between 2013 and 2018. During this period, the
SEEC monitored the implementation of efciency
measures in existing and new plants (SEEC 2018).
We do not assume that this trend continues at
the same level in the EE-Mod variant. The Saudi
industrial base, whose energy mix is dominated
by oil and gas, already has the lowest electricity
intensity among the G20 countries.15
We assume that the electricity intensities of PRV,
PUB and AGR decrease by 23% between 2018
and 2030 (i.e., 2.2% per year). This efciency gain
corresponds to the savings estimated by Krarti et
al. (2017) for Saudi buildings, including commercial
and governmental buildings. They estimate these
savings based on actions with no signicant costs,
such as thermostat adjustments or replacements of
existing lighting. Finally, we assume that residential
electricity consumption decreases by 10%
between 2018 and 2030 (i.e., 0.9% per year). This
assumption is based on the cost-free gains in Saudi
residential buildings estimated by Krarti et al. (2017).
14
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
4. Electricity demand scenarios through 2030
In the EE-High variant, we assume that the
observed 2013-2018 efciency gains for IND of 2.4%
per year continue through 2030. For PRV, PUB and
AGR, we assume that intensity falls by 50% from
2018 to 2030 (i.e., 5.6% per year). This assumption
corresponds to the Level-3 efciency improvements
estimated by Krarti et al. (2017).16 Finally, we assume
that residential electricity demand decreases by
26% from 2018 to 2030 (i.e., 2.5% per year). This
assumption is based on the higher efciency gains
estimated by Krarti et al. (2017). Table 3 summarizes
the intensity assumptions of the EE scenarios.
Table 3. Electricity intensities of the energy efciency scenarios (index = 1 in 2013).
a Scenarios of residential sector are for the sector’s total consumption.
Source: Authors’ estimates based on ECRA (2018), Enerdata (2021) and SAMA (2019) for 2013 and 2018. Authors’ assumptions for
projected values.
Variant Sector 2013 2018 2025 2030 ∆ 2018-2030
EE-Mod Residentiala1.00 1.14 1.08 1.03 -10%
Industrial 1.00 0.85 0.72 0.64 -25%
Commercial 1.00 1.23 1.05 0.94 -24%
Government 1.00 1.25 1.07 0.95 -24%
Agriculture 1.00 1.00 0.85 0.76 -24%
EE-High Residentiala1.00 1.14 0.96 0.84 -26%
Industrial 1.00 0.85 0.57 0.43 -50%
Commercial 1.00 1.23 0.84 0.64 -48%
Government 1.00 1.25 0.85 0.65 -48%
Agriculture 1.00 1.00 0.68 0.52 -48%
Finally, we highlight one important aspect of this
scenario’s narrative. We model efciency only
variants to assess the potential energy savings that
can be achieved under various intensity targets.
Although efciency measures may incur some
costs, such as investments in new appliances
or retrotting, we consider these costs to be
insignicant. The prevailing electricity intensity
in Saudi Arabia is high compared with other
countries. This comparison may be biased for
industrial consumption, as the Kingdom’s power
mix is dominated by fossil fuels. Nevertheless,
Saudi electricity demand can be signicantly
reduced through costless or very-low-cost policies
with relative immediate feasibility. Such policies
may include awareness campaigns, updates to
regulations to improve codes and standards for
new buildings, and the introduction of dynamic
pricing (Faruqui et al. 2011). Moreover, the current
deployment of smart meters in the Kingdom will
help with tracking and reducing inefcient electricity
use.17 Finally, the mid-term projection horizon
suggests that energy intensity reductions may be
achievable through demand-side management
measures, as we assume in the efciency variants.
15
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
5. Results
We begin our discussion of the results
by presenting the electricity demand
in the REF scenario. In this scenario,
total electricity demand reaches 365.4 TWh in
2030, up from 299.2 TWh in 2018. In the period
without statistical data, that is, from 2019 to 2030
(see section 3.2), projected demand growth is
signicantly below its historical trend. Between 2009
and 2018, total electricity demand grew at 5.3% per
year on average. In contrast, we project average
electricity demand growth of 1.6% per year between
2019 and 2030.
Before discussing the long-term outcomes by
sector, we highlight two observations related to the
aggregate results for the early projected years. First,
the modeled demand declines by 1.5% in 2019. This
result occurs because we assume that the electricity
intensities for the IND, PRV and PUB sectors
maintain their 2013-2018 averages throughout the
projection period (see section 4.1). This nding does
not have signicant implications for the relevance
of our results. Indeed, preliminary data for 2019
indicate that total Saudi electricity sales fell by about
1% (Soummane 2020).
Second, we consider the effects of the 2020
COVID-19 pandemic and the associated lockdown
measures on electricity demand. In many regions
worldwide, electricity demand dropped signicantly.
However, Saudi electricity consumption data
suggest that demand remained roughly stable
in the Kingdom. The higher contributions of the
residential sector and warmer weather may have
offset the impacts of the pandemic (Soummane and
Peerbocus 2020).
We now consider the sectoral breakdown of demand
in the REF scenario. The residential sector remains
the primary consumer of electricity. However, its
share drops from 43.6% of the total electricity
demand in 2018 to 39.0% in 2030 in this scenario.
The two EPR rounds signicantly impacted
residential demand. In 2016, residential demand
started to atten for the rst time. It then declined
by 9.1% in 2018. The average growth rate between
2009 and 2018 therefore dropped to 3.2%.18 Under
our assumption that real prices remain at 2018
levels, residential demand slowly recovers at an
average rate of 1.1% per year between 2019 and
2030, ending at 142.4 TWh.
In the other sectors, electricity demand grows
faster than in the residential sector. Indeed, keeping
electricity intensities stable results in electricity
demand increasing in line with the sectors’ outputs.
IND’s output grows at a rate of 1.8% through 2030,
and its electricity demand increases from 58.2 TWh
in 2018 to 81.9 TWh in 2030. IND comprises 22.4%
of the total electricity demand in 2030 in the REF
scenario, a 3.0 percentage point (pp) increase from
its share in 2018.
In the PRV sector, electricity demand reaches
82.1 TWh in 2030, up from 61.8 TWh in 2018. This
sector’s electricity demand is projected to grow at
a rate of 1.9%. It accounts for 22.5% of the total
electricity demand in the REF scenario in 2030,
a 6.8 pp increase over 2018. In the PUB sector,
electricity demand reaches 51.5 TWh in 2030, up
from 43.9 TWh in 2018. The annual growth rate of
electricity demand in this sector between 2019 and
2030 is 2.1%, which is in line with projected output
growth. This sector’s demand accounts for 14.2%
of the total electricity demand in the REF scenario
in 2030. This share reects a decrease from 19.7%
in 2018, but it is similar to PUB’s 2013 share of total
demand. PUB’s average electricity intensity over
2013-2018 is 12.8% higher than the 2013 calibration
level. If PUB maintains this electricity intensity over
the projected period, electricity demand in this
sector grows slightly faster than in the IND and
16
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
5. Results
PRV sectors. Finally, electricity demand in AGR is
projected to increase at an average annual rate of
1.7%. It is expected to reach 7.4 TWh by 2030, up
from 4.9 TWh in 2018. This sector’s demand is only
2.0% of total demand in 2030, up slightly from 1.6%
in 2018 (Figure 2).
Figure 2. Electricity demand by sector in the REF scenario.
Sources: ECRA (2018) for 2013-2018 data. IMACLIM-SAU for projected values.
50
100
150
200
250
300
350
400
2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030
TWh
Residential Industrial Commercial Government Agriculture
projection
historical
Next, we analyze the alternative demand scenarios.
Figure 3 illustrates the impacts of price reforms and
efciency measures on electricity demand. Aligning
domestic electricity prices with those of emerging
G20 countries (PR-EM variant) reduces annual
demand growth by half compared with the REF
scenario. Specically, demand growth in the PR-EM
variant is 0.7% per year between 2019 and 2030,
compared with 1.6% per year for REF. In 2030, total
demand in the PR-EM variant is 40.5 TWh less
than that in the REF scenario (i.e., 11.1% lower).
Further tariff increases to meet average G20 price
levels (PR-AVG variant) result in roughly stable total
demand over the projection horizon. The growth rate
is -0.1% per year on average through 2030. Total
demand in this scenario is 71.6 TWh less than in the
REF scenario (i.e., 19.6% lower).
In the EE scenarios, efforts to reduce electricity
intensity in the IND, PRV and PUB sectors, along
with reduced residential demand, drive higher
energy savings. By 2030, achieving the objectives of
the EE-Mod variant attens electricity demand over
the projection horizon. In 2030, aggregate electricity
17
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
demand is 45.6 TWh lower in the EE-Mod variant
than in the REF scenario (i.e., 12.5% lower). The
larger efciency gains of the EE-High variant
reduce total electricity demand by 2.2% per year on
average. This decrease is equivalent to around 6
TWh per year through 2030. In 2030, total demand
is 118.7 TWh lower than in the REF scenario (i.e.,
32.5% lower).
5. Results
Figure 3. Total electricity demand by scenario.
Sources: ECRA (2018) for 2013 and 2018. IMACLIM-SAU results for projected values.
262.7
299.2
365.4
327.6
300.4 319.8
246.6
0
50
100
150
200
250
300
350
400
2013 2018 REF G20-EM G20-AVG EE-Mod EE-High
TWh
2030
Energy Efficiency
Price Reforms
Residential Industrial Commercial Government Agriculture
The electricity consumption patterns in the two
alternative scenarios are similar to those in the
REF scenario. The residential segment remains the
largest consumer of electricity in all scenarios. In the
PR-EM and PR-AVG variants, residential demand
accounts for 40.8% and 41.5% of total demand in
2030, respectively. By comparison, it accounts for
39.0% of total demand in the REF scenario. The
shares of most of the other sectors in the PR-EM
and PR-AVG variants are similar to those of the REF
scenario. The exception is PUB’s demand, which
decreases slightly from 14.4% in the REF scenario
to 12.9% and 12.3% in the PR-EM and PR-AVG
variants, respectively.
In the EE-Mod and EE-High variants, residential
demand in 2030 accounts for 40.4% and 43.1% of
total demand, respectively. Moreover, the shares of
IND in total demand in 2030 are 21.7% and 24.7%
in the EE-Mod and EE-High variants, respectively.
These shares are similar to those in the REF
scenario. The shares of the remaining sectors in
total demand are roughly similar for the EE-Mod
variant and the REF scenario. They are slightly
18
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
lower in the EE-High variant than in the REF
scenario.
The PR-EM and PR-AVG variants achieve
electricity savings of 40.5 TWh and 71.6 TWh,
respectively, in 2030 (Figure 4). In the PR-EM
variant, the main electricity-consuming sectors
reduce their consumption by around 10 TWh
each. AGR, whose consumption is relatively
marginal, contributes savings of 1.6 TWh. In
this variant, the residential sector reduces its
consumption by 7.0% relative to REF by 2030.
IND and PRV reduce their relative consumption
by around 12%, and PUB’s relative
consumption falls by 18.4%. In the PR-AVG
variant, residential demand is 14.3% less than
in the REF scenario in 2030, amounting to
savings of 20.4 TWh. IND and PRV contribute
16.2 TWh and 17.3 TWh, respectively, to
demand savings. These declines correspond to
abatements of 19.8% and 21.1%, respectively,
relative to the REF scenario in 2030.
Figure 4. Sectoral differences in the PR and EE scenarios relative to the REF scenario in 2030.
Sources: IMACLIM-SAU simulation results.
-25
-20
-15
-10
-5
0
Residential Industrial Commercial Government Agriculture
TWh
G20-EM G20-AVG
-50
-40
-30
-20
-10
0
Residential Industrial Commercial Government Agriculture
TWh
EE-Mod EE-High
The electricity demand savings of the EE scenarios
are greater than those of the PR scenarios. Relative
to the REF scenario, electricity demand is 45.6
TWh and 118.7 TWh lower in 2030 under the
EE-Mod and EE-High variants, respectively. The
residential sector accounts for declines of 13.1 TWh
and 36.1 TWh, respectively. The efciency gains
for intermediate users result in signicant demand
abatement. Consumption in the IND sector is 12.5
TWh (-15.3%) and 21.1 TWh (-25.7%) lower under
the EE-Mod and EE-High scenarios, respectively,
than under the REF scenario in 2030. Likewise,
demand in the PRV and PUB sectors is 11.2 TWh
and 7.2 TWh lower under the EE-Mod variant,
respectively. In both sectors, demand is about 14%
lower than in the REF scenario. In the EE-High
variant, the PRV and PUB sectors achieve electricity
savings of 35.6 TWh and 22.3 TWh, respectively.
Demand in these sectors is about 43% lower in this
scenario than in the REF scenario.
5. Results
19
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
Table 4 reports the main macroeconomic indicators
for our scenarios. The overall outcomes of our
alternative scenarios are positive compared with
the REF scenario. Indeed, real GDP is higher in all
four variants than it is in the REF scenario, and it
is the highest in the EE scenario. The differences
in real GDP across our scenarios stem from the
contrasting effects of lower electricity demand on
households’ revenues and the public budget.
In the PR scenario, price hikes to meet
international benchmarks negatively impact
households’ purchasing power and, ultimately, their
real revenue. In 2030, households’ real revenue is
0.6% (SAR 10.0 billion) lower in the PR-EM variant
than in the REF scenario. The corresponding gure
for PR-AVG is 3.2% (SAR 50.4 billion). However,
the price increases in the PR scenario positively
impact the public balance because they alleviate
the burden of government subsidies. In 2030, the
public budget balance (as a share of GDP) is 1.7
pp higher in the PR-EM variant than in the REF
scenario. It is 4.1 pp higher in the PR-AVG variant
than in the REF scenario. The accumulated budget
balance gains for PR-EM and PR-AVG yield net
debt gains (as a share of GDP) of 16.3 pp and 41.9
pp, respectively, in 2030 compared with the REF
scenario.
Modeling electricity demand in a general equilibrium
framework allows us to compute the implications
of simulated scenarios from a macroeconomic
perspective. Ultimately, the overall positive impacts
on GDP of 0.72% in PR-EM and 0.75% in PR-AVG
are mainly driven by higher public spending. The
government’s scal outlook is better in these
scenarios than in the REF scenario. However,
the deterioration of household revenue in these
scenarios translates into weaker domestic demand.
Thus, although the net effect on GDP is positive, the
unemployment rate is higher in the PR scenario than
in the REF scenario. The unemployment rates in the
PR-EM and PR-AVG variants are 0.2 pp and 1.1 pp
higher than in the REF scenario, respectively. These
numbers correspond to 31,000 and 172,000 jobs
lost, respectively.
Table 4. Macroeconomic indicators by scenario.
Source: IMACLIM-SAU simulations. pp = percentage points.
Difference from REF in 2030
Variable 2013 REF 2030 PR-EM PR-AVG EE-Mod EE-High
Real GDP
(billions of 2013 riyals) 2,773.3 4,177.2 +0.7% +0.8% +0.9% +2.5%
Unemployment rate 5.6% 7.4% +0.2 pp +1.1 pp -0.3 pp -0.8 pp
Trade balance
(% of GDP) 24.6% 8.7% +0.4 pp +1.2 pp +0.1 pp +0.2 pp
Government budget
balance (% of GDP) 8.8% 2.9% +1.7 pp +4.1 pp +0.3 pp +0.7 pp
Net public debt
(% of GDP) -95.9% -108.0% -16.3 pp -41.9 pp -2.0 pp -4.8 pp
5. Results
20
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
As mentioned above, the EE scenario yields higher
gains from a macroeconomic perspective. Indeed,
fostering efciency benets both household revenue
and the public budget. Reducing the intensity of
electricity use lowers overall electricity consumption
in the targeted sectors, thereby decreasing public
spending on nancial incentives. The government
balance improves by 0.3 pp and 0.7 pp in the
EE-Mod and EE-High variants, respectively. Net
debt as a share of GDP is 2.0 pp and 4.8 pp lower in
the EE-Mod and EE-High variants than in the REF
scenario, respectively.
Moreover, the EE-Mod variant reduces the revenue
loss relative to the PR scenarios, and the EE-High
variant provides some revenue gain. In 2030,
households’ real revenue differs by -0.4% and
+0.8% under the EE-Mod and EE-High variants,
respectively, compared with the REF scenario.
Domestic consumption is consequently higher in
the EE scenarios, resulting in improved employment
outlooks. In 2030, the unemployment rates are 0.3
pp and 0.8 pp lower under the EE-Mod and EE-High
variants, respectively, compared with the REF
scenario. These numbers correspond to 42,000 and
126,000 jobs created, respectively. In 2030, GDP is
0.9% and 2.5% higher in the EE-Mod and EE-High
variants, respectively, than in the REF scenario. This
result is driven by a combination of a positive scal
outlook and greater household revenue.
5. Results
21
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
6. Conclusions
The Saudi electricity sector is undergoing
structural changes. For decades, it
grew rapidly, supported by government
incentives through regulated low prices for
electricity consumption. These demand trajectories
were deemed unsustainable, threatening the
government’s scal sustainability and crowding
out valuable fossil fuel exports. In recent years,
the authorities have launched ambitious programs
to curb demand growth and reduce wasteful uses
of electricity. These public action plans have not
only reformed prices but also promoted efciency
measures.
Given these structural changes, this study presents
trajectories for future Saudi electricity demand
through 2030. We discuss options for reducing
electricity demand in different sectors while
maintaining economic growth. The results from our
pricing reform and energy efciency scenarios can
provide Saudi policymakers with insights about the
potential outcomes of different policies.
We modify the dynamic CGE model, IMACLIM-SAU
(Soummane et al. 2022) to reect the features of the
electricity sector. We then use this model to explore
three future power demand scenarios. Our reference
scenario contains no additional electricity price
reforms and no enforcement of efciency measures.
In our price reform scenario variants, we simulate
regulated prices converging to the average prices
for emerging G20 countries and all G20 countries.
Finally, we run two variants of an energy efciency
scenario in which electricity intensity efciency
improvements are enforced, assuming either
moderate or high gains.
In our reference scenario, demand growth is
signicantly below its historical trend, at 1.6% per
year on average between 2019 and 2030. Demand
reaches 365.4 TWh by 2030, up from 299.2 TWh in
2018. Our alternative scenarios show the potential
for signicant savings relative to the reference
scenario. Aligning electricity prices with those
of emerging G20 countries reduces aggregate
demand by 40.5 TWh (11.1%) in 2030. Reaching
the average price of all G20 countries abates total
electricity demand by 71.6 TWh (19.6%) in the same
year. However, opting for efciency measures to
rationalize electricity use may provide even greater
savings. In the moderate energy efciency variant,
total electricity demand may be 45.6 TWh (12.5%)
lower than in the reference scenario. This difference
may reach 118.7 TWh (32.5%) under the ambitious
energy efciency targets of the high energy
efciency variant.
We also analyze the potential macroeconomic
effects of our scenarios. In the four alternative
variants, real GDP improves overall in 2030
compared with the reference scenario. This outcome
stems from the favorable effects of price reforms
and efciency measures on the public budget.
Lower electricity demand improves the government’s
scal balance because it alleviates the nancial
burden associated with power generation subsidies.
The impacts on the public budget and debt are
greater under the price reform scenario than under
the energy efciency scenario. However, the gains
in the former scenario are slightly undermined by the
decreases in households’ real revenue and producer
competitiveness due to higher electricity prices.
One important driver of electricity demand that
our analyses do not explicitly account for is future
large-scale projects. Indeed, although our model
captures sectoral interdependencies, it does not
incorporate the government’s announcements
of several large projects being established
across Saudi Arabia. These projects include the
development of the tourism industry, with expanded
Hajj capacity, and several industrial initiatives (e.g.,
22
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
the National Industrial Development and
Logistics Program). These projects may generate
signicant incremental electricity demand over
the coming decade. This study focuses on
projecting structural sectoral electricity demand
through 2030. Future research should account
for these large projects and extend our projection
horizon.
6. Conclusions
23
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
Endnotes
1 Pumped-storage hydropower is a large-volume electricity storage solution. However, we do not
consider this option further in this analysis. See Matar and Shabaneh (2020) for details on the potential of
pumped-storage hydropower in the Kingdom.
2 This gure (i.e., 299.2 TWh) corresponds to billed nal consumption. It does not include in-plant power
use or transmission and distribution losses.
3 He also models aggregate electricity demand in other Gulf Cooperation Council (GCC) countries (i.e.,
Bahrain, Kuwait, Oman, Qatar and the United Arab Emirates).
4 Eltony and Mohammad (1993) model sectoral electricity demand in aggregate across GCC countries.
5 Atalla et al. (2018) compare total and non-oil GDP as proxies for income. The two proxies lead to
signicantly different estimates of the income elasticity of Saudi demand for gasoline of 0.09 and 0.61,
respectively.
6 Le Treut (2017) models households’ consumption of energy goods as the sum of exogenous basic needs
and price-elastic uses. This distinction lacks a usable assessment in the case of Saudi Arabia.
7 Soummane et al. (2022) treat electricity production and the three other types of energy production as
exogenous.
8 Because estimated elasticities of substitution for Saudi production are lacking, we use published
estimates for other countries (see Annex A). We analyze the sensitivity of our results to these key
parameters and to other socioeconomic variables in Annex B.
9 Industry corresponds to the SAMA (2019) sectors of Other Mining and Quarrying Activities;
Manufacturing; and Electricity, Gas and Water. Private services correspond to Construction; Wholesale
& Retail Trade, Restaurants and Hotels; Transport, Storage & Communication; Finance, Insurance,
Real Estate & Business Services; and Community, Social & Personal Services. The government sector
corresponds to Producers of Government Services.
10 The choice of specication does not signicantly impact projected demand. For instance, the projected
demand in the REF scenario under Specication 1 deviates from the projected demand under Specication
2 by 1.8% in 2030 (365.4 TWh versus 359.0 TWh).
11 The sample of G20 emerging countries consists of Argentina, Brazil, China, India, Indonesia, Mexico,
Russia, Saudi Arabia, South Africa and Turkey. The G20 also includes Australia, Canada, France,
Germany, Italy, Japan, South Korea, the United Kingdom and the U.S. The electricity prices for the selected
countries are derived from the Enerdata (2021) database.
24
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
Endnotes
12 Although our price targets are similar to those of Hasanov (2019), he assumes that the targets are
reached in 2025.
13 Given the scope of this study, we focus on energy efciency measures on the demand side. Large
energy savings in terms of oil and gas feedstocks can also be achieved by switching to efcient or
alternative generation units (Alyousef and Abu-Ebid 2012; Matar et al. 2017).
14 The SEC’s consumption rationalization measures are available at https://www.se.com.sa/en-us/Pages/
IndustrialSector.aspx for industry, https://www.se.com.sa/en-us/Pages/CommercialSector.aspx for the
commercial sector and https://www.se.com.sa/en-us/Pages/GovernmentalSector.aspx for the government
sector.
15 We compute electricity intensity as the ratio of value-added from industry (in real U.S. dollars) to the
electricity consumption of the sector (in gigawatthours) using data from Enerdata (2021).
16 In addition to the measures cited for the EE-Mod variant, the main measure underpinning these gains is
the replacement of air conditioning units.
17 https://www.se.com.sa/en-us/customers/Pages/SmartMeters.aspx
18 Between 1999 and 2008, residential electricity demand grew at an average rate of 6.6% per year.
25
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
Al-Faris, Abdul Razak. 2002. “The Demand for
Electricity in the GCC Countries.ˮ Energy Policy
30(2):117–24. DOI: 10.1016/s0301-4215(01)00064-7
Al-Hawwas, Abdullah. 2010. “Policy Experiments for
the Saudi Economy Using a Computable General
Equilibrium Model (CGE): Oil Demand and Tariff
Liberalisation Effects on the Saudi Economy.ˮ PhD
diss., University of Dundee.
Al-Sahlawi, Mohammed. 1990. “Forecasting
the Demand for Electricity in Saudi Arabia.ˮ
The Energy Journal 11(1):119–25. DOI: 10.5547/
issn0195-6574-ej-vol11-no1-10
Al-Thumairi, Imtithal A. 2012. “Fixed Exchange
Rate or Flexible Exchange Rate for Saudi Arabia:
Optimal Solution of CGE Model.ˮ International
Journal of Basic and Applied Sciences 1(2):67–87.
DOI: 10.14419/ijbas.v1i2.29
Aldubyan, Mohammad, and Anwar Gasim. 2021.
“Energy Price Reform in Saudi Arabia: Modeling
the Economic and Environmental Impacts and
Understanding the Demand Response.ˮ Energy
Policy 148:111941. DOI: 10.1016/j.enpol.2020.111941
Alkhareif, Ryadh. M., William A. Barnett, and Nayef
Alsadoun. 2017. “Estimating the Output Gap for
Saudi Arabia.ˮ International Journal of Economics
and Finance 9(3):81–90. DOI: 10.5539/ijef.v9n3p81
Alyousef, Y., and M. Abu-Ebid. 2012. “Energy
Efficiency Initiatives for Saudi Arabia on Supply
and Demand Sides.ˮ In Energy Efficiency-A Bridge
to Low Carbon Economy, edited by Zoran Morvaj,
297–308. IntechOpen. DOI: https://10.5772/38660
Anouti, Y., S. Elborai, R. Kombargi and H. Hage
(2020). Electricity Pricing Reform: A Bitter Pill
for GCC Industries, Strategy&. https://www.
strategyand.pwc.com/m1/en/reports/2020/electricity-
pricing-reform/electricity-pricing-reform.pdf
Arab Petroleum Investments Corpopration
(APICORP) 2018. “Saudi Energy Price Reform
Getting Serious.ˮ APICORP Energy Research
3(5):1–3.
Atalla, Tarek N., Anwar A. Gasim and Lester C.
Hunt. 2018. “Gasoline Demand, Pricing Policy,
and Social Welfare in Saudi Arabia: A Quantitative
Analysis.ˮ Energy Policy 114:123–33. DOI: 10.1016/j.
enpol.2017.11.047
Atalla, Tarek N., and Lester C. Hunt. 2016.
“Modelling Residential Electricity Demand in the
GCC Countries.ˮ Energy Economics 59:149–58.
DOI: 10.1016/j.eneco.2016.07.027
Beckman, Jayson, Thomas Hertel, and Wallace
Tyner. 2011. “Validating Energy-oriented CGE
Models.ˮ Energy Economics 33(5):799–806. DOI:
10.1016/j.eneco.2011.01.005
Billmeier, Andreas, and Dalia Hakura. 2008.
“Trade Elasticities in the Middle East and Central
Asia: What is the Role of Oil?ˮ International
Monetary Fund Working Papers 2008(216). DOI:
10.5089/9781451870749.001
Böhringer, Christoph, and Thomas F. Rutherford.
2013. “Transition Towards a Low Carbon Economy:
A Computable General Equilibrium Analysis for
Poland.ˮ Energy Policy 55:16–26. DOI: 10.1016/j.
enpol.2012.11.056
References
26
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
References
Cai, Yiyong, and Vipin Arora. 2015. “Disaggregating
Electricity Generation Technologies in CGE
Models: A Revised Technology Bundle Approach
with an Application to the U.S. Clean Power
Plan.ˮ Applied Energy 154:543–55. DOI: 10.1016/j.
apenergy.2015.05.041
Chi, Yuanying, Zhengquan Guo, Yuhua Zheng,
and Xingping Zhang. 2014. “Scenarios Analysis
of the Energiesʼ Consumption and Carbon
Emissions in China Based on a Dynamic CGE
Model.ˮ Sustainability 6(2):487–512. DOI: 10.3390/
su6020487
De Santis, Roberto A. 2003. “Crude Oil Price
Fluctuations and Saudi Arabia's Behaviour.ˮ
Energy Economics 25(2):155–73. DOI: 10.1016/
s0140-9883(02)00106-8
Devarajan, Shantayanan, and Sherman Robinson.
2002. “The Impact of Computable General
Equilibrium Models on Policy.ˮ Presented at a
conference on “Frontiers in Applied General
Equilibrium Modelingˮ sponsored by the Cowles
Foundation and held at Yale University, New Haven,
CT, April.
Diabi, Ali. 1998. “The Demand for Electric Energy
in Saudi Arabia: An Empirical Investigation.ˮ OPEC
Review 22(1):13–29. DOI: 10.1111/1468-0076.00039
Eltony, M. Nagy, and Yousuf H. Mohammad. 1993.
“The Structure of Demand for Electricity in the Gulf
Cooperation Council Countries.ˮ The Journal of
Energy and Development 18(2):213–21.
Enerdata. 2020. “Energy & CO2 Data.ˮ https://
www.enerdata.net/expertise/energy-co2-data.html
Accessed on March 10, 2020.
Energy and Cogeneration Regulatory Authority
(ECRA). 2018. Annual Statistical Booklet for
Electricity and Seawater Desalination Industries.
Riyadh: ECRA.
Faruqui, Ahmad, Ryan Hledik, Greg Wikler,
Debyani Ghosh, Joe Prijyanonda, and Nilesh Dayal.
2011. “Bringing Demand-side Management to
the Kingdom of Saudi Arabia.ˮ The Brattle Group
2011:40 – 41.
Gelan, Ayele. 2018. “Economic and Environmental
Impacts of Electricity Subsidy Reform in Kuwait:
A General Equilibrium Analysis.ˮ Energy Policy
112:381–98. DOI: http://doi.org/10.1016/j.
enpol.2017.10.032
Hasanov, Fakhri 2019. “Theoretical Framework
for Industrial Electricity Consumption Revisited:
Empirical Analysis and Projections for Saudi Arabia.ˮ
KAPSARC Discussion Paper. Riyadh: KAPSARC.
DOI: 10.30573/ks--2019-dp66
Hasanov, Fakhri, Frederick L. Joutz, Jeyhun
Mikayilov, and Muhammad Javid. 2020. “KGEMM:
A Macroeconometric Model for Saudi Arabia ˮ
KAPSARC Discussion Paper. Riyadh: KAPSARC.
DOI: 10.30573/ks--2020-dp04
He, Y. X., L. F. Yang, H. Y. He, T. Luo, and Y. J.
Wang. 2011. “Electricity Demand Price Elasticity in
China Based on Computable General Equilibrium
Model Analysis.ˮ Energy 36(2):1115–23. DOI:
10.1016/j.energy.2010.11.038
Holmøy, Erling. 2005. “The Anatomy of Electricity
Demand: A CGE Decomposition for Norway.ˮ
Discussion Papers No. 426, Statistics Norway,
Research Department.
International Energy Agency (IEA). 2019. World
Energy Outlook 2019. Paris: IEA.
27
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
References
International Monetary Fund (IMF). 2016.
World Economic Outlook Database, April 2016.
Washington, D.C.: International Monetary Fund.
Johansen, Leif. 1960. A Multi-Sectoral Study of
Economic Growth. Amsterdam: North Holland
Publishing Co.
Kat, Bora, Sergey Paltsev, and Mei Yuan. 2018.
“Turkish Energy Sector Development and the Paris
Agreement Goals: A CGE Model Assessment.ˮ
Energy Policy 122:84–96. DOI: 10.1016/j.
enpol.2018.07.030
Kingdom of Saudi Arabia. 2017. Fiscal Balance
Program Balanced Budget 2020. Riyadh: Kingdom
of Saudi Arabia. https://vision2030.gov.sa/sites/
default/files/attachments/Fiscal%20Balance%20
Program%202017.pdf
———. 2019. Fiscal Balance Program 2019 Update.
Riyadh: Kingdom of Saudi Arabia. https://www.
mof.gov.sa/en/budget/2019/Documents/Fiscal%20
Balance%20Program.pdf
Koesler, Simon, and Michael Schymura.
2015. “Substitution Elasticities in a Constant
Elasticity of Substitution Framework–Empirical
Estimates Using Nonlinear Least Squares.ˮ
Economic Systems Research 27(1):101–21. DOI:
10.1080/09535314.2014.926266
Krarti, Moncef, Kankana Dubey, and Nicholas
Howarth. 2017. “Evaluation of Building Energy
Efficiency Investment Options for the Kingdom of
Saudi Arabia.ˮ Energy 134:595–610. DOI: 10.1016/j.
energy.2017.05.084
Le Treut, Gaëlle. 2017. “Methodological Proposal for
Hybrid Modelling: Consequences for Climate Policy
Analysis in an Open Economy (France).ˮ Ph.D.
dissertation, École Doctorale Ville, Transports et
Territoires (VTT)-ED 528.
Lin, Boqiang, and Zhujun Jiang. 2011. “Estimates
of Energy Subsidies in China and Impact of Energy
Subsidy Reform.ˮ Energy Economics 33(2):273–83.
DOI: 10.1016/j.eneco.2010.07.005
Liu, Wei, and Hong Li. 2011. “Improving Energy
Consumption Structure: A Comprehensive
Assessment of Fossil Energy Subsidies Reform
in China.ˮ Energy Policy 39(7):4134–43. DOI:
10.1016/j.enpol.2011.04.013
Matar, Walid, Frederic Murphy, Axel Pierru,
Bertrand Rioux, and David Wogan. 2017.
“Efficient Industrial Energy Use: The First Step in
Transitioning Saudi Arabia's Energy Mix.ˮ Energy
Policy 105:80–92. DOI: 10.1016/j.enpol.2017.02.029
Matar, Walid, and Rami Shabaneh. 2020. “Viability
of Seasonal Natural Gas Storage in the Saudi
Energy System.ˮ Energy Strategy Reviews
32:100549. DOI: 10.1016/j.esr.2020.100549
Mikayilov, Jeyhun, Frederick L. Joutz, and Fakhri
Hasanov. 2019. “Gasoline Demand in Saudi Arabia:
Are the Price and Income Elasticities Constant.ˮ
KAPSARC Discussion Paper. Riyadh: KAPSARC.
DOI: 10.30573/ks--2019-dp81
Mikayilov, Jeyhun I., Abdulelah Darandary,
Ryan Alyamani, Fakhri J. Hasanov, and Hatem
Alatawi. 2020. “Regional Heterogeneous Drivers
of Electricity Demand in Saudi Arabia: Modeling
Regional Residential Electricity Demand.ˮ Energy
Policy 146:111796. DOI: 10.1016/j.enpol.2020.111796
28
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
Nachet, Said, and Marie-Claire Aoun. 2015. “The
Saudi Electricity Sector: Pressing Issues and
Challenges.ˮ IFRI Security Studies Center.
Saudi Arabian Monetary Authority (SAMA). 2019.
Annual Statistics 2019. Riyadh: SAMA. http://www.
sama.gov.sa/en-US/EconomicReports/Pages/
YearlyStatistics.aspx
Saudi Energy Efciency Center (SEEC). 2018.
SEEC Annual Report. Riyadh: SEEC. https://
www.seec.gov.sa/en/media-center/annual-report/
annual-report-for-2018/
Shehabi, Manal. 2020. “Diversication Effects
of Energy Subsidy Reform in Oil Exporters:
Illustrations from Kuwait.ˮ Energy Policy 138:110966.
DOI: 10.1016/j.enpol.2019.110966
Soummane, Salaheddine. 2021. “Overview
of Saudi Electricity Demand (1970–2018).ˮ
KS--2021-CO03, KAPSARC. https://
www.kapsarc.org/research/publications/
overview-of-saudi-electricity-demand-1970-2018/
Soummane, S., Frédéric Ghersi and Franck
Lecocq (2022). “Structural Transformation
Options of the Saudi Economy Under Constraint
of Depressed World Oil Prices.ˮ The Energy
Journal 43(3): 181-200. DOI: http://doi.
org/10.5547/01956574.43.3.ssou
Soummane, Salaheddine, Frédéric Ghersi, and
Julian Lefèvre 2019. “Macroeconomic Pathways
of the Saudi Economy: The Challenge of Global
Mitigation Action Versus The Opportunity of National
Energy Reforms.ˮ Energy Policy 130:263–82. DOI:
10.1016/j.enpol.2019.03.062
Soummane, Salaheddine, and Nawaz Peerbocus.
2020. “Impact of the COVID-19 Lockdown on
Electricity Demand: Global Insights and Implications
for Saudi Arabia - Part I.ˮ KS--2020-II19. Riyadh:
KAPSARC. https://www.kapsarc.org/research/
publications/impact-of-the-covid-19-lockdown-on-
electricity-demand-global-insights-and-implications-
for-saudi-arabia/
United Nations Department of Economic and
Social Affairs (U.N. DESA). 2015. World Population
Prospects: The 2015 Revision. New York: United
Nations.
Wing, Ian Sue. 2006. “The Synthesis of Bottom-up
and Top-down Approaches to Climate Policy
Modeling: Electric Power Technologies and the
Cost of Limiting US CO2 Emissions.ˮ Energy Policy
34(18):3847–69. DOI: 10.1016/j.enpol.2005.08.027
References
29
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
In this annex, we present the core socioeconomic
drivers common to our three simulated scenarios,
REF, PR and EE (Table A.1). In Annex B, we
analyze the sensitivity of our results to variations in
these parameters.
Annex A. Model parametrization
Table A.1. Socioeconomic drivers common to all scenarios.
Driver Unit 2013 2030
Labor endowment Million full-time equivalents 19.9 27.6
Labor productivity Index equal to 1 in 2013 1.00 1.10
Default export trend Index equal to 1 in 2013 1.00 1.76
Gross xed capital formation Share of GDP 23.9 26.1
Oil output Million barrels per day 9.2 12.9
Oil price U.S. dollars per barrel 115 88
The primary drivers of economic growth are the
exogenous labor endowment and productivity
trajectories. We measure labor endowment by
the population in the 15-65 years age group,
derived from the United Nations (UN 2015). For
labor productivity, we use annual data reported by
Oxford Economics between 2013 and 2017. After
2017, we assume that labor productivity grows at
1.3% per year through 2030. This rate corresponds
to the average productivity growth for the Saudi
economy between 2011 and 2015 estimated by
Alkhareif et al. (2017).
We apply the default export trend to
non-energy goods to define the baseline
subject to terms-of-trade variations.
The growth of Saudi export markets is
approximated by that projected by the
International Monetary Fund (IMF 2016) for
the Middle East and North Africa region.19
Investment, that is, gross fixed capital
formation, is a constrained share of GDP
calibrated based on SAMA (2019) data
between 2013 and 2017. It is assumed to
converge to its 2013-2017 average in 2030.
The oil sector plays an important role in the
Saudi economy. The two key variables driving its
performance, that is, output and price, are both
exogenous in IMACLIM-SAU. Their trajectories are
based on the International Energy Agency’s (IEA’s)
Stated Policies Scenario Saudi oil output reaches
12.9 million barrels per day in 2030 (IEA 2019). The
international oil price is projected to recover from its
low 2015 level to $81 per barrel in 2025 and $88 per
barrel in 2030 (IEA 2019).
19 See Equation A-30 of Soummane et al. (2022).
30
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
Annex A. Model parametrization
The production functions described in section 3.1
build on the elasticities of substitution and trade
elasticities presented in Table A.2. More detailed
algebraic descriptions are provided in Annex A of
Soummane et al. (2022).
Table A.2. Substitution and trade elasticities.
σVA σXσM
OIL 1.28 - -
GAS 1.28 - -
REF 0.31 - -
ELE 0.30 - -
AGR 0.34 0.67 -0.09
MIN 1.28 0.67 -0.09
CHM 0.52 0.67 -0.09
NMM 0.29 0.67 -0.09
MAN 0.25 0.67 -0.09
PRV 0.27 0.67 -0.09
PUB 0.43 0.67 -0.09
TRA 0.87 0.67 -0.09
OTP 0.56 0.67 -0.09
Sources: Koesler and Schymura (2015), Billmeier and Hakura (2008).
At the bottom of the production function, capital,
labor and electricity are treated as substitutable
inputs to the production of value-added (VA).
Moreover, non-energy goods are elastic to the terms
of trade, with elasticities of σX for exports and σM for
imports. More details are provided in Annex A of
Soummane et al. (2022).
31
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
In this section, we present a sensitivity analysis
of electricity demand to the socioeconomic
variables listed in Annex A. For brevity, we
present the sensitivity results for aggregate demand
in the REF scenario and the variation in GDP at the
end of the projection horizon. The main goal of this
analysis is to show how variations in the selected
model drivers may impact electricity demand in
2030. The macroeconomic and social drivers in the
model are subject to uncertainties. Thus, we test
them within reasonable boundaries corresponding
to “low” and “high” variants (Table B.1).
Annex B. Sensitivity analysis
Table B.1. Exogenous socioeconomic variables in the model in 2030.
Variable Unit Low REF High
Active population Million full-time equivalents 26.2 27.6 29.0
Labor productivity Index equal to 1 in 2013 1.05 1.10 1.27
Default export trend Index equal to 1 in 2013 1.41 1.76 2.12
Gross xed capital formation Share of GDP 23.9 26.1 29.8
Elasticity of substitution, σVA n.a. 0.5* Reference value Reference value 1.5* Reference value
Oil output Million barrels per day 11.0 12.9 13.5
Oil price U.S. dollars per barrel 62 88 111
We vary the active population endowment by
-/+5% of the REF scenario target in 2030. For
labor productivity, we assume a moderate gain
of 5% by 2030 in the low variant, corresponding
to the targeted increase from Oxford Economics.
We assume a gain of 27% in the high variant,
corresponding to the non-oil productivity gains
estimated by Alkhareif et al. (2017). We vary the
default export trend by -/+20% of the REF scenario
target in 2030. We interpret the low export variant
as weaker growth in the Middle East and North
Africa as the region fails to recover economically
from the COVID-19 pandemic. The high variant
reflects a quick post-pandemic recovery of the
region’s economies.
No proper estimates of elasticities of substitution
(σVA) in the Saudi context are available. Thus, we
also conduct sensitivity analyses for the parameters
presented in Table A.2. To do so, we multiply these
elasticities by factors of 0.5 (low variant) and 1.5
(high variant), respectively. For gross fixed capital
formation, investment returns to its 2013 level in the
low variant. In the high variant, investment reaches
its 2015 level. In 2015, the government stimulated
the non-oil sector with public spending in reaction to
an oil price slump. We use the percentage variations
in oil production in the Middle East region projected
by the IEA (2019). Production in the low and high
variants corresponds to the IEA’s Sustainable
Development Scenario (SDS) and Current Policies
Scenario (CPS), respectively.20 Likewise, the low
20 The IEA (2019) does not project Saudi Arabia’s oil production in its alternative CPS and SDS scenarios.
32
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
Annex B. Sensitivity analysis
and high variants for projected oil prices correspond
to oil prices in the IEA (2019) SDS and CPS.
The analysis reveals that electricity demand is more
sensitive to variations in investment, oil prices and
productivity levels. It is moderately sensitive to
variations in the active population (Figure B.1). Oil
production levels and the default export trend for
non-energy goods only marginally impact electricity
demand. The variants used for the presented
sensitivity analysis are based on the REF scenario.
In other words, they have xed electricity prices
at 2018 levels and xed electricity intensities of
production at the 2013-2018 average. The exception
is the sensitivity analysis for σVA, which is based
on production Specication 1. Thus, the variations
in electricity demand stem from impacts on
households’ revenue dynamics, sectoral outputs and
the scal balance. These variables, along with other
macroeconomic variables (e.g., the unemployment
rate, trade balance and public debt) drive the
changes in electricity demand trajectories. Together,
these variations result in changes in GDP.
Figure B.1. Sensitivity of total electricity demand to macroeconomic variables, difference from the REF
scenario in 2030.
Source: IMACLIM-SAU simulations.
-10.0 -5.0 0.0 5.0 10.0
Investment
Sigma_VA
Oil price
Productivity
Active population
Export trend
Oil production
TWh
Low High
As stated previously, we restrict the sensitivity
analysis for macroeconomic variables to induced
variations in real GDP relative to its 2030 level in
the REF scenario. As expected, higher variation
in real GDP is associated with higher variation in
total electricity demand, with the exception of oil
production. GDP in 2030 varies by -2.1% and +3.4%
relative to the REF scenario in the low and high
investment variants, respectively. Oil prices in the
low and high variants are associated with GDP
33
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
Annex B. Sensitivity analysis
changes of -2.3% and +1.8%, respectively. In the
low variants for the active population, productivity
and the default export trend, GDP changes by
-0.8%, -0.6% and -0.3%, respectively. In the high
variants for those factors, it changes by +0.5%,
+1.3% and +0.3%, respectively, compared to the
REF scenario.
Finally, reducing the elasticity of substitution of the
inputs to electrified value added is associated with
higher electricity consumption but reduces GDP
by 0.5%. Increasing this elasticity of substitution
improves GDP by 0.2%. Indeed, lowering the
substitutability of primary production factors
increases the share of electricity, as it becomes
a rigid production factor with fewer substitutes.
However, because electricity has a lower share of
value added, increasing its share compared with
the other factors slightly reduces GDP. Conversely,
increasing substitutability at the bottom of the
production function allows competing factors (i.e.,
labor and capital) to substitute for electricity in the
production process. However, this substitutability
warrants additional bottom-up expertise, as further
electrifying some industrial processes requires
retrofits or upgrades to production mechanisms.
34
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
About the Authors
Salaheddine Soummane
Salaheddine is a research associate in the Energy Transitions and Electric Power
program. His current research focus includes modeling the Saudi electricity market,
including its reforms and regulatory framework.
Prior to joining KAPSARC, Salaheddine worked as a research associate for the Centre
for International Research on Environment and Development (CIRED), a National
Center for Scientic Research (CNRS) lab, based in Paris, where he focused on
integrated economy-energy modeling. He has also worked as an economist researcher
within the research and development (R&D) department of the utility group EDF
(Paris). He was part of the group’s energy markets and environmental regulation unit
focused on emerging markets.
Salaheddine holds a Ph.D. in Economics from Paris-Saclay University (France), an
M.Sc. in Energy Economics from the University of Montpellier (France), and an M.Sc.
in Finance from the Aix-Marseille School of Economics (France).
Frédéric Ghersi
Frédéric Ghersi is a French CNRS researcher assigned to the Centre International
de Recherche sur l’environnement et le Développement (CIRED), Paris. Since
1997, he has been working on hybrid bottom-up and top-down modeling of
economy-energy-environment (3E) interactions, which he applies to develop
outlooks on the efciency and equity of climate policies and energy transitions.
His current research focuses on the macroeconomics of climate nance, on the
distributional impacts of the French National Low Carbon Strategy (SNBC), and on
the transfer of the IMACLIM method (model and data hybridization) to international
academic partners in the framework of the IMACLIM Network. The network now
extends to 11 countries, including France and the BRICS states (Brazil, Russia, India,
China, and South Africa).
35
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
About the Project
The electricity sector is at the heart of energy transition in many countries. Both the demand and
the supply side of the electricity market require rigorous assessments to identify the appropriate
policy options that can yield the maximum benet for stakeholders. This project focuses on the
demand side of the Saudi market. The supply side, including the evolution of the power mix and
the integration of the domestic electricity market with regional markets, is assessed in other
projects.
Electricity demand in Saudi Arabia has grown consistently over the past decades. Since 2016,
Saudi Arabia has initiated price reforms and launched rationalization campaigns for energy use,
to curb electricity demand and inefcient use. As a result, Saudi electricity demand attened
between 2016 and 2018, eventually dropping for the rst time on record in 2019. The various
factors determining potential demand growth or decline and the impact of energy efciency on
electricity demand growth in developing economies are not well understood.
Understanding electricity demand growth is critical for public policy development. Uncertainty
regarding electricity demand growth rates directly impacts investment needs. This project
disentangles the primary driving factors of Saudi electricity demand. We analyze potential future
trends of electricity demand by sector in Saudi Arabia.
36
Projecting Saudi Sectoral Electricity Demand in 2030 Using a Computable General Equilibrium Model
www.kapsarc.org
