Energy 290 (2024) 130109

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Integrated hybrid multi-regional input-output for assessing life cycle air

emissions of the Italian power system
Le Quyen Luu a, c, *, Thomas Gibon b, Maurizio Cellura a, Eleonora Riva Sanseverino a,
Sonia Longo a
a
University of Palermo, Department of Engineering, Viale delle Scienze Bld.9, 90128, Palermo, Italy
b
Luxembourg Institute of Science and Technology, Environmental Research and Innovation, Maison de l’Innovation, 5 Avenue des Hauts-Fourneaux, L-4362, Esch-sur-
Alzette, Luxembourg
c
Vietnam Academy of Science and Technology, Institute of Energy Science, A9 Building, 18 Hoang Quoc Viet, Cau Giay, Viet Nam

A R T I C L E I N F O A B S T R A C T

Keywords: The air emissions of the Italian power system, as well as national emissions between 2010 and 2017 and pro­
Hybrid multi-regional input-output jections to 2040, have been assessed from a lifecycle perspective, using an integrated hybrid two-region input-
Power system output model of Italy versus the rest of the world. The Italian economy is divided into 42 sectors, including
Scenario integrated
electricity, which is further disaggregated into seven technologies. Detailed electricity sector data, from Istat, are
Life cycle thinking
Air emissions
fed into the EXIOBASE input-output database. NAMEA tables represent overall air emissions, while the Ecoinvent
Production and consumption-based accounting database is used for the electricity sector. Electricity transition scenarios from Terna and Snam have been in­
tegrated into input-output and air emission databases. Demand and emissions were tracked within the electricity
sector over medium-term, and the findings showed a sharp decrease between 2017 and 2025, from 97.5 MtCO2
to 32.6 MtCO2. By 2040, air emissions from the electricity sector are expected to grow gradually, compared to
those of 2030, from 22.2 MtCO2 to 25.9 MtCO2, suggesting that the demand between 2030 and 2040 grows faster
than the decarbonization effort during the same period. There is an overall, gradual downtrend between 2010
and 2040, with all air emission categories declining by half from both production and consumption-based per­
spectives in this period.

capable of evaluating the path of decarbonization on a large scale, and to

1. Introduction take into account both the energy sector and other economic sectors.
Moreover, the GHG reductions of the renewable energy technology
Rising temperatures and other adverse consequences of climate adoption should be considered with a life cycle thinking approach,
change require a comprehensive socio-economic transition, especially in extending beyond the operation stage of the technologies to encompass
the energy sector, due to its significant contribution to global green­ their entire life cycle.
house gas (GHG) emissions. There is a strong global energy transition The energy sector constitutes three-quarters of global GHG emissions
toward renewable energy sources (RES), improving energy efficiency [4,5]. The heat and electricity sector alone contributed about 42 % of
and a process of electrification, concerning both the supply of energy the energy sector’s GHG emissions globally [6]. Besides, the energy
and the demand sectors of industries, building and transportation sec­ sector is a significant source of nitrogen oxides (NOx), sulphur oxides
tors. Several energy transition scenarios have been developed by the (SOx) and particulate matter (PM) [7]. Due to the large contribution of
International Renewable Energy Agency on energy production and the energy sector particularly electricity generation, to the total GHG
consumption [1], and the International Energy Agency on energy in and other air emissions, these sectors should be prioritized to reduce
buildings [2]. global emissions.
The purpose of the energy transition scenarios aim at decarbonizing Electricity sector air emissions should be considered under both
the energy sector and keeping global temperature rise below 2 ◦ C, which production (PBA) and consumption-based accounting (CBA) perspec­
align with the goal of carbon neutrality by 2050, as outlined in the tives with the life cycle thinking approach. Current GHG reporting sys­
European Green Deal [3]. These scenarios require hybrid tools which are tems, such as those of United Nations Framework Convention on Climate

* Corresponding author. University of Palermo, Department of Engineering, Viale delle Scienze Bld.9, 90128, Palermo, Italy.
E-mail address: [email protected] (L.Q. Luu).

https://doi.org/10.1016/j.energy.2023.130109
Received 9 April 2023; Received in revised form 23 October 2023; Accepted 21 December 2023
Available online 28 December 2023
0360-5442/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
L.Q. Luu et al. Energy 290 (2024) 130109

Nomenclature N2O Nitrous oxides

NACE Classification of economic activities of European
CBA Consumption-based accounting Community
CH4 Methane NAMEA-Air National Accounting Matrix with Environmental
CO2 Carbon dioxide Accounts Air Emissions
CSP Concentrating solar power NMVOC Non-methane volatile organic compound
EEIOA environmentally extended input-output analysis NOx Nitrogen oxides
GHG Greenhouse gas PBA Production-based accounting
H-MRIO Hybrid multi-regional input-output analysis PFC Perfluorocarbon
HFC Hydrofluorocarbon PM Particulate matter
IOA Input-output analysis RES Renewable energy resource
IOT Input-output table RoW Rest of the world
LCA Life cycle assessment SF6 Sulphur hexafluoride
LCI Life cycle inventory SOx Sulphur oxides
MRIO Multi-regional input-output analysis

Change and Convention on Long-range Transboundary Air Pollution, the process-based inventory data can provide product-level details, but
require countries to report on their annual GHGs as well as other air may introduce truncation errors due to the exclusion of some processes
emissions, using a PBA perspective and following the so-called “terri­ or system incompleteness. For example, a case study on copper wire
torial” principle. This accounting principle records emissions based on indicates that the system boundary gap of LCA may cause up to 60 % of
their physical sources, arising from the territory of a given country truncation error [14]. This occurs because some processes are excluded
regardless of who emits. Its limitation is the omission of emissions during different supply chain tiers [15]. On the other hand, EEIOA
embodied in trade, import and export. This raises the question of presents an economic-wide view and does not require allocation prac­
emission responsibility, borne either by the producing or the consuming tice, but may cause aggregation errors during the grouping of several
entities. Both perspectives are useful in policymaking, and may lead to products into a single sector. The complexity of the supply chain causes
distinct, yet complementary decisions. For example, CBA covers blind challenges in identifying an appropriate system boundary, which sug­
spots of PBA, such as carbon leakage phenomena in highly-connected gests the need for an integrated hybrid approach for GHG and air
countries. On the methodological side, however, complementing PBA emission accounting [16].
with CBA drives the need for a more diverse accounting method. IOA (and EEIOA) traces economic (and environmental) flows within
This study quantifies and assesses air emissions of Italy, focusing on a nation or a region, without considering international trade. Multi-
the Italian electricity sector projections from both PBA and CBA per­ regional input-output (MRIO) analysis extends the boundary to
spectives between 2010 and 2040. By considering life cycle air emissions include several regions or nations, incorporating import and export
of the Italian electricity sector, three specific aspects will be examined: flows among them [17,18]. This feature allows addressing questions
(1) what makes up of the changes in Italian air emission in the period related to countries’ responsibilities. For example, a country that is
2010–2017 and projections to 2040; (2) how the air emissions of the highly dependent on imported resources, that are more energy- or
Italian electricity sector are impacted by other economic sectors and vice emission-intensive than their local counterparts, would not be incen­
versa; and (3) what are the relations between the air emissions tivized to develop local industry if only territorial emissions are
embodied in trade and consumption activities, and their implications on accounted for – even though it would decrease global emissions.
air emission reduction efforts. Another question is how to effectively reduce energy consumption or
This study contributes to the methodological aspects of environ­ mitigate emissions at the international level. Therefore, instead of hy­
mental accounting by filling the literature gap of computing sectorial bridizing LCA and IOA, the hybridization is conducted on LCA and
and national air emissions from both PBA and CBA perspectives with a MRIO, which will be called “H-MRIO” in this paper.
life cycle thinking approach, and helps to guide policy decisions. At the Integrated hybrid analysis has been applied for quantifying the life
same time, the study provides a case study on the Italian electricity cycle impacts of electricity generation technologies as well as in other
transition scenarios to examine the changes in sectorial and national air sectors such as energy consumption in buildings [19], construction
emissions between 2010 and 2017, and projections to 2040; and identify materials [20], transportation [21], national import [22] and cities [23].
the interactions among air emissions of different economic sectors, in In the electricity sector, Gibon et al. used the H-MRIO approach to
relations with imports and exports. integrate energy scenarios in an MRIO table to assess the life cycle GHG
Methods for life cycle GHG and air emission accounting include emissions of concentrating solar power (CSP) [24]. The model is a
process-based life cycle assessment (LCA) [8], environmentally combination of process-based life cycle inventory (LCI) and MRIO tables
extended input-output analysis (EEIOA) [9] and hybrid approach of LCA with forecasted technological and resource changes up to 2050. It was
and input-output analysis (IOA) [10]. EEIOA was developed as an found that life cycle GHG emissions of CSP varied between 33 and 95
additional layer to economic IOA, the principal objective of which is gCO2e/kWh (depending on specific regions) in 2010 and would reduce
quantifying direct and indirect input requirements to meet a given final to 30–87 gCO2e/kWh in 2050.
demand. The product system is thereby considered from a top-down Similarly, a H-MRIO model was proposed in China by Li et al. to
point of view, starting from the whole economic supply chain, and evaluate the life cycle CO2 emissions, energy consumption and energy
narrowing it down into economic sectors and product groups. In EEIOA, payback time of a 10-MW CSP plant [25]. Vélez-Henao and Vivanco
both direct and indirect emissions of the consumed product groups are conducted a H-MRIO study on wind power plants, using on-site data for
quantified and compared among different categories of household direct emissions and supply chain data for indirect emissions [26].
consumption, investment or export [11,12]. Different electricity technologies and various life cycle impacts were
The choice of different LCA modelling approaches, with different extended in Hertwich et al.’s study, which investigated the co-benefits
sources of data and allocation practice, is a source of uncertainty during and trade-offs of decarbonizing the electricity sector in terms of life
the quantification and assessment of the product system [13]. LCA with cycle GHG and non-GHG impacts [27]. At global scale, Wan et al.

2
L.Q. Luu et al. Energy 290 (2024) 130109

studied the direct and indirect impacts of water consumption from (3) In combination with the power development scenarios, the Ital­
power mix adjustment in the world largest seven emitting economies ian electricity sector is disaggregated into seven electricity gen­
[28]. In de Koning’s study, a MRIO database (EXIOBASE) was used to eration technologies, for both intermediate flow matrices and
calculate CO2 emissions of production activities to meet final demand by final demand vectors in Italian IOT. Similarly, in the stressor
2050 [29]. The author developed several scenarios taking into account matrices, the air emissions of the electricity sector are dis­
socio-economic development, efficiency and technology improvement, aggregated into those of seven electricity generation technolo­
low-carbon energy technologies and shifts in production and con­ gies, with data taken from Ecoinvent. At this time, the H-MRIO
sumption. The study concluded that it is difficult to reach the 2 ◦ C model consists of 42 sectors (36 economic sectors - 1 electricity
reduction targets relying on low-carbon energy technologies alone. sector + 7 power technologies).
It is worth noting that the number of H-MRIO studies remains (4) The model is calculated with historical data of 2010 and 2017,
somewhat limited. Existing literatures applied H-MRIO either on indi­ and replicated for the future scenarios of 2025, 2030 and 2040.
vidual power technologies or on an electricity mix. The studies con­
ducted on electricity mix, e.g. Refs. [27–29] considered the global The hybridization and MRIO calculation procedures are written in
emissions as a whole. Therefore, this study aims to contribute to H-MRIO the Python language, provided in the Supplementary Information. The
literature on the electricity sector by integrating power development mathematical framework and specific data for H-MRIO analysis are
scenarios and analysing the sectorial and national emissions from a presented below.
global perspective. Main points of the existing literature applied inte­
grated H-MRIO in the energy sector are presented in Table 1.
2.2. H-MRIO mathematical framework
This study utilized the H-MRIO approach for quantifying and
assessing the life cycle GHG and other air emissions of Italy and its
The calculations of life cycle requirements, as well as associated
electricity sector, taking into account of the electricity transition sce­
impacts (here, air emissions) are carried out following Leontief equa­
narios. The national and sectoral air emissions are considered from both
tions (1)–(4) for IOA and EEIOA.
PBA and CBA perspectives, including imports to and exports from Italy
Production-consumption balance
and their relationship with the air emissions from the rest of the world
(RoW). The ultimate goal of the study is to support energy, economic Ax + y = x (1)
and environmental policies in reducing the national air emissions from
Solving the balance equation
power consumption and production, considering trade dynamics, and
avoiding transferring the impacts of one sector or country to another. x = (I − A)− 1 y = Ly (2)
Calculating (direct) stressor coefficients
2. Methodology
S = F̂x − 1
(3)
2.1. Conceptual framework
Calculating life cycle multipliers
The H-MRIO is applied following the under-described framework M = SL (4)
(illustrated in Fig. 1).
In which:
(1) First, two types of data, including MRIO and hybridization data x is the vector of total gross outputs of the economy needed to meet
are collected. MRIO data such as the Italian and multi-regional the final demand, dimension n*1.
input-output tables (IOTs) and air emissions accounts are I is the identity matrix, A is the intermediate flow coefficient matrix,
collected from Istat and EXIOBASE. Hybridization data is dimension n*n.
collected from Italian electricity/energy suppliers [30] for power L is the Leontief matrix, dimension n*n
development scenarios, and from the Ecoinvent database for y is the vector of final demand of products, dimension n*1.
direct air emissions of electricity generation technologies. S is the stressor coefficients matrix, dimension m*n.
(2) From MRIO data, the MRIO model with two regions of Italy and F is the stressor matrix, dimension m*n.
RoW and 36 economic sectors will be constructed. M is the multiplier matrix.
In a multi-regional context, the life cycle emissions of the PBA and
i i
CBA of region i, Fpba and Fcba can be expressed in equations (4) and (5):
Table 1 i
Fpba = Fsi e + Fyi (4a)
Main points of the existing literature applied integrated H-MRIO in the energy
sector. i
Fcba = Myi + Fyi (5)
Paper Studied product Geographical Impacts
boundary In which:
i
[24] One technology Global Life cycle GHG emissions Fpba is the life cycle emission of the production-based accounts of
(CSP)
region i
[25] One technology Country (China) Life cycle CO2 emissions, energy
(CSP) consumption and energy payback Fsi is the direct emission from production activities of region i, and e
time an appropriate summation vector
[26] One technology Country Several life cycle environmental Fyi is the direct emission from consumption activities of region i
(Wind) (Columbia) impacts, including climate
i
change, eutrophication, Fcba is the life cycle emission of the consumption-based accounts of
acidification, toxicity, etc. region i
[27] Several Global Various life cycle impacts,
M and yi are the multiplier matrix and the vector of final demand of
electricity including GHG and non-GHG
technologies impacts region i
[28] Global electricity Global Direct and indirect impacts of With the inclusion of power development scenarios, the H-MRIO
mix water consumption model is adapted following the THEMIS model [24]. THEMIS model
[29] Low carbon Global CO2 emissions provided the global perspective, while this H-MRIO estimated and
technologies
assessed the national emissions from domestic, import and export

3
L.Q. Luu et al. Energy 290 (2024) 130109

Fig. 1. Main principles of the disaggregation, and integration of energy scenarios.

activities. The hybrid process is implemented for each matrix and vector, 2.3. Data
including the technical coefficient matrix A, final demand vector y and
stressor matrix F. An example of hybridization of matrix A for the energy 2.3.1. Intermediate flow coefficient matrix
sector is presented in equation (6): The intermediate flow coefficient matrix A presents the relationship
among different industries (or sectors) of the economy, in which prod­
At− ele = ̂v Tt− es ∗ A2017− ele ∗ vt− es (6) ucts (or outputs) of one industry are used as inputs of other industries, or
In which: in other words, they indicate the inter-industrial relations of the amount
At− ele is the disaggregated matrix of the electricity sector in year t of intermediate products to produce other products. The matrix A is
A2017− ele is the original vector of the electricity sector in 2017 developed based on the Italian IOT for the year 2017 [31]. For the
vt− es is the vector of electricity share in year t matrix A in future scenarios, the data is taken from IOT with the inte­
̂ T
v t− es is the transpose vector of ves, ̂
v is the diagonal matrix with the gration of power development scenarios by 2025, 2030 and 2040. The
Italian IOTs are published every five years by Istat, with a breakdown of
principal diagonal elements being taken from vector v
the economy into 63 industrial sectors, corresponding to 63 products.
Further information on the hybridization equations for each matrix
The sectors and products are classified by activity, with reference to the
and vector can be seen in the supplementary information, Fig. 2 for
classification of economic activities of European Community (NACE).
hybridizing the matrix A, the vector y, and Fig. 3 for hybridizing the
The coding structure of CPA corresponds to that of NACE up to the
matrix F.
fourth level [32].

Fig. 2. Hybridization of matrix A and vector y.

4
L.Q. Luu et al. Energy 290 (2024) 130109

Fig. 3. Hybridization of matrix F.

2.3.2. Power development scenarios The details of final consumption for each economic sector are estimated
This section describes the Italian electricity generation technologies based on the average share of final consumption during the period
in 2017 and its development pathway by 2040. According to the data of 2015–2017. Besides, it is essential to disaggregate the final demand from
the National Trend Italy developed by Terna and Snam [30], the Italian the electricity sector into final demand of specific electricity generation
power technologies in 2017 comprised of (1) natural gas-based power, technology, using the production share of the technologies.
(2) coal and other fossil fuel-based power, (3) hydropower, (4) wind
power, (5) solar power, and (6) other RES-based power. By 2025–2040, 2.3.4. Stressor coefficient matrix
the generation mix will change towards decreasing of fossil fuels and The stressor coefficient matrix S indicate the stressor F per total gross
increasing of RES. These scenarios of Snam and Terna are adapted to the output x of the economy. The data set for the matrix F in 2017 was taken
electricity sector for the hybridization of the matrix A. Fig. 4 below in­ from the National Accounting Matrix with Environmental Accounts Air
dicates the percentage of six electricity generation technologies along Emissions (NAMEA-Air) of Italy. The Italian NAMEA-Air tables are
with net import/export (which is considered as the seventh technology) published annually by Istat [35], presenting 10 atmospheric emission
contributing to the grid mix between 2010 and 2040. categories, namely CO2, CH4, N2O, SOx, NOx, NH3, CO, NMVOC, PM2.5
This power development scenario will be integrated into the Italian and PM10 for 63 products (production sectors) and three household
IOT. However, in the Italian IOT, the electricity, gas and steam sector are consumption activities of transport, heating and others [35].
presented as one single product, encompassing three subsectors: elec­ The NAMEA-Air shares the same framework and similar classifica­
tricity supply, gas supply, and steam and air conditioning supply, not tion with IOT, with 63 industries/sectors. In NAMEA-Air, the emissions
limited to electricity generation alone. Among these subsectors, elec­ of the electricity sector are not further divided into specific electricity
tricity generation represents 82.4 % of the sectoral value added in 2017 generation technologies. Therefore, they are disaggregated into seven
[33]. Therefore, it is assumed that electricity generation technologies electricity generation technologies using the data set from Ecoinvent for
can be the representative of the electricity, gas, and steam sector. air emissions of electricity generation technologies [36]. It should be
noted that the emissions of Ecoinvent data is more diverse than that of
2.3.3. Final demand vector NAMEA-Air. Therefore, the two databases need to be matched. The
IOTs present the link between the inputs used and outputs manu­ matching rules follow the description on the coverage of substances of
factured by the production sectors, and the final demand (y vector). In NAMEA-Air [37].
this study, data on final demand, extracted from the Italian IOT for 2017 Matrix F of future years (F2025, F2030, F2040) is forecasted using linear
are utilized for vector y2017. The total final demand of 2025, 2030 and regression. In this case, the emission volume of the economy undergoes a
2040 (y2025, y2030, y2040) is forecasted based on the total final con­ gradual change during the period 2017–2040 due to economic devel­
sumption data of Organisation for Economic Co-operation and De­ opment, productivity, and efficiency improvement.
velopment’s studies estimated for 2023 [34], using linear regression.

Fig. 4. National trend electricity production scenario for Italy, 2010–2040, fossils are indicated in bold, imports in italics [30].

5
L.Q. Luu et al. Energy 290 (2024) 130109

2.3.5. Aggregation of economic sectors With the change in final demand and electricity sector composition
To construct the 2-region MRIO table, data for the first region, Italy is of Italy, consumption-based GHG emissions appear to decrease in the
extracted from Istat, and RoW is added as the second region, with data period 2010–2040 (Fig. 5). Specifically, due to changes in production
taken from EXIOBASE database. Instead of using EXIOBASE directly for structure, emission coefficients, and final demand, the annual CO2
both Italy and RoW, the study used Istat data for Italy and EXIOBASE for emission reduction embodied in production activities during the period
RoW. This decision increases the workload of matching two databases. 2017–2025 will be up to 7.1 MtCO2, which makes up 57.1 MtCO2
EXIOBASE is a multi-regional environmentally extended supply-use emission reduction in the whole period. The increased final demand of
table and IOT available for the years 1995–2011, covering 49 coun­ Italy causes an annual increase of 4.8 MtCO2. The change in the pro­
tries and regions (28 EU members, 16 major economies and five RoW duction structure, including the electricity sector and corresponding
regions). The IOTs are built upon the supply-use table of 163 industries change in other economic sectors, helps to reduce 6.1 MtCO2 annually.
and 200 products, illustrating the structural change by integrating The change in emission flow coefficients brings an annual reduction
economic development as reported by national statistics agencies. The credit of about 5.8 MtCO2. During the period of 2025–2030 and
data is detailed for energy, agricultural production, resource extraction 2030–2040, the annual change in emission reduction will be much
and bilateral trade [38]. Until now, the database is now-casted to 2022 smaller, at 2.3 MtCO2 and 33.9 ktCO2, respectively.
and the most updated version is EXIOBASE 3.8.2 (Stadler et al., 2021).
The classification of industries and products in the EXIOBASE is 3.2. Change in air emissions of Italian electricity sector
based on NACE1. The classification of industries and product in Istat
data (for both IOTs and air emission accounts) were based on NACE1 The hybridization of power development scenarios causes a change
until 2010. After 2010, the classification changed into NACE2. Besides, in all air emissions categories in 2017, at various scales. The smallest
the number of Istat economic sectors is 63 and that of EXIOBASE is 200 difference occurs in CO2, at an 8 % difference. The largest difference
(product by product). The mismatch in sector classification (and num­ occurs in PM2.5, which is followed by SOx. The difference in other air
ber) between the Italian and RoW databases requires a common emissions: N2O, CH4, NOx, NH3, NMVOC, and PM10 ranges between
concordance to match in these two databases’ economic sectors. In this − 0.98 and 2.28. This difference is mainly caused by the level of aggre­
study, the economy is firstly aggregated into 36 production sectors and gation of process-based LCI and input-output data. Input-output data is
later disaggregated into 42 sectors. The aggregation and disaggregation taken from NAMEA, which reports air emissions of the electricity, gas
of economic sectors in different databases can be seen in the supple­ and steam sector, while LCI data includes air emissions of electricity
mentary information. generation technologies. First, this mismatch causes the omission of
emissions from gas supply and steam and air supply in the LCI data used
3. Results and discussion in this study. The production of natural gas is a CH4-intensive process
[43]. Therefore, the missing of emissions from gas supply in the LCI data
3.1. Decomposition of Italian consumption-based emissions will omit an amount of CH4 emissions from this subsector, which ex­
plains the lower CH4 emissions of hybrid results compared to the orig­
The total GHG emissions to meet global final demand in 2017 inal NAMEA. Second, the air emissions of electricity generation
calculated with this model is 47.69 GtCO2e. This number is slightly technologies in LCI data are gathered for the seven ‘representative’
higher than the global GHG emissions estimated by Climate Watch, at 47 technologies contributing to most of the electricity generation. In
GtCO2e excluding land use change and forestation [4]. The difference in practice, the number of ‘actual’ technologies goes beyond seven. The
the obtained results of this model and other models was caused by the emissions are not the same for ‘representative’ and ‘actual’ technologies,
difference in the scope of air emissions being studied. This model has which causes a difference between the hybrid results and the original
been developed based on Istat database for Italy and EXIOBASE data for input-output data.
RoW. Both Istat and EXIOBASE databases are actual anthropogenic After 2017, due to the change in electricity generation technologies
emissions of CO2, CH4 and N2O, excluding emissions from land use land and power consumption, the future air emissions dramatically reduce in
use change and forestation, and biomass burning as a fuel. Meanwhile, the electricity sector, as presented in Fig. 6. Most of the PBA emissions of
the Climate Watch’s model takes into account all GHGs (CO2, CH4, N2O, the electricity sector come from fossil fuel-based electricity, e.g., coal
and F-gases such as HFCs, PFCs, and SF6), excluding land use change and and natural gas. A smaller part comes from other RES, including
forestation. This causes a difference of around 1 GtCO2e of F-gases and geothermal and biomass-based electricity. The production of solar and
2.8 GtCO2e of CH4. Moreover, Climate Watch’s model excludes wind power does not generate any airborne emission, and that of hy­
short-cycle biomass burning such as agricultural waste burning and dropower emits N2O only. The reduction in electricity from fossil fuels
savanna burning, but includes other biomass burning such as forest fires, such as coal and natural gas contributes to reducing the PBA emissions of
post-burn decay, peat fires and decay of drained peatlands. The exclu­ this sector nearly four-fold from 97.5 MtCO2 in 2017 to 25.9 MtCO2 by
sion of emissions from land use (mostly CH4), biogenic CO2 and F-gases 2040. With regards to CBA, CO2 emissions total 34.9 MtCO2 in 2017, and
in this model leads to an insignificant difference of around 0.69 GtCO2e then drop by more than half, at 13.7 MtCO2 by 2040. The CBA CO2
(less than 1.5 %). emission of the electricity sector is divided among technologies by their
Another calculation with this model for GHGs from combustion ac­ production structure. As it can be observed, low-carbon technologies
tivities only, and excluding fugitive emissions of CH4, GHG emissions such as solar and wind power technologies contribute to emissions,
from agriculture, waste management and industrial production, indi­ because of the manufacturing of their infrastructures. The CBA emis­
cated that the GHG emissions to meet global final demand in 2017 is at sions of electricity are smaller than the PBA ones, as they are shared by
33.96 GtCO2e. In this case, the GHG emissions from combustion account other economic sectors as intermediates for production activities. CBA
up to 70 % of the total GHG emissions. This result is 3.1 % higher than takes into account the emissions of electricity as a final product, and
the reported number of the International Energy Agency on CO2 emis­ excludes the emissions of electricity as an intermediate for other pro­
sions for the energy sector in 2017, at 32.92 GtCO2e [39,40]. duction activities.
In order to look into details of the sources of the change in the air
emission, a decomposition analysis has been conducted following [41]. 3.3. Change in emissions of other Italian economic sectors by years
A similar study on air emission change in Italian household consumption
between 1999 and 2006 [42] shows that, between 1999 and 2006, the The absolute change in air emissions of the electricity sector induces
indirect CO2 emission from Italian household consumption was about 13 a change in the economy emission structure, as presented in Fig. 7. In
MtCO2. 2017, from PBA perspective, the electricity sector accounts for the

6
L.Q. Luu et al. Energy 290 (2024) 130109

Fig. 5. Decomposition of the Italian consumption-based emission variation over four periods. Note: the periods are not of equal duration, variations are indicated as
the average annual variation within each period.

largest share of the national emission of CO2 (36 %), which reduces to change in their CO2 emission shares such as rubber and plastics, water
18 % by 2040. At the same time, the CO2 emission shares of agriculture, transportation, coke and petroleum and land transportation (though
forestry and fishing, wholesale and retail trade, water and waste man­ their large contribution to the total emissions), reduce by 1% point
agement, and air transportation increase about 2–5% points each be­ during the same period. Apart from the electricity sector, other eco­
tween 2017 and 2040. Some economic sectors, which have the smaller nomic sectors see a considerable change in their emissions (absolute

7
L.Q. Luu et al. Energy 290 (2024) 130109

Fig. 6. Air emissions of the Italian electricity sector by year. Top: all air emissions, bottom left: production-based CO2 emissions, bottom right: consumption-based
CO2 emissions.

Fig. 7. Contribution of economic sectors to life cycle emissions (excluding import, export and trade).

value) during the period 2017–2040. For example, the CO2 emissions of 2017, which reduces to 5.9 % by 2040. The CO2 emission shares of some
land transportation reduce from 15.64 MtCO2 in 2017 to 10.57 MtCO2 other economic sectors also decrease during the period 2017–2040, such
by 2040, 32 % in the whole period or 1.4 % annually. as construction and healthcare (reducing around 1% point). Meanwhile,
From CBA perspective, the changes in electricity consumption the CO2 emission shares of some sectors increase, such as food and
induce changes in other economic sectors, which are clearly shown in beverage (increasing less than 1% point). It should be noted that the CO2
coke and petroleum, pharmaceuticals, water transportation, education, emission contributions of these sectors to the national CBA emissions do
and healthcare, either increasing or decreasing their emissions. Partic­ not show the correspondingly absolute increase (or decrease). Instead,
ularly, electricity sector accounts for 11.6 % of the total CO2 emissions in they relatively present the changes in the identified “hotspot” sectors

8
L.Q. Luu et al. Energy 290 (2024) 130109

over years. beverage, electricity and construction (refer back to Fig. 7). In 2017,
The absolute values of the CO2 emissions decrease in all economic wholesale and retail contribute to more than 12 % of the total CBA CO2
sectors between 2017 and 2040. The decrease is clearly presented in emission of Italy. The four remaining sectors account for an average CBA
economic sectors such as construction, decreasing from 20.99 MtCO2 in CO2 emission, from 6 % to 10 % of the total CO2 emissions. By 2040, the
2017 to 13.4 MtCO2 by 2040, at about 0.33 MtCO2 annually; or food and shares of emissions of these sectors remain in the same range. This
beverage, decreasing from 15 MtCO2 to 12.5 MtCO2, or 0.1 MtCO2 emission pattern suggests that between 2017 and 2040, in order to
annually; or healthcare, decreasing from 17.7 MtCO2 to 11.43 MtCO2 or reduce the national CO2 emissions, effort should be focused on these
0.27 MtCO2 annually in the same period (see Fig. 7). ‘hotspot’ sectors. Besides, the different contributions of domestic and
import emissions to the total emissions suggest that Italy should have
proper strategies to reduce its emissions in terms of geographical effort.
3.4. Share of emissions embodied in import out of consumption
CO2 emissions of Italian trade partners for food and beverage, health,
construction, and wholesale and retail should be taken into account
Emissions from imports account for a significant share of the national
because their emissions largely depend on imports (see Fig. 8). The effort
CBA emissions in most of air emission categories. Specifically, CO2
should be taken either to reduce their trade partners’ emission intensity,
emissions embodied in imports hold up to 43.9 % of CBA emissions. This
or to move away from trade partners that have high emission intensities.
indicates the outsourcing of Italian air emissions, as well as its emission
Meanwhile, equal effort should be shared between local manufacturers
dependency on foreign products. In order to reduce the emissions of
and trade partners being relevant to renewable power technologies such
Italy, it is necessary to take into account of imported products, and
as solar, wind and other renewables.
emissions of its trade partners. Emissions from some economic sectors
are more dependent on those of imported products than others, which is
3.5. Uncertainty analysis and sensitivity analysis
expressed by the close or loose relation between air emissions embodied
in import and consumption in these economic sectors, as presented in
The uncertainty of the CO2 emissions of the Italian electricity sector
Fig. 8. Some economic sectors with large shares of air emissions
is analysed using Monte Carlo simulation for 1000 runs. In 2017, the
embodied in imports compared to those of consumption include trade,
production-based CO2 emissions of the Italian electricity sector range
pharmaceutical, computer and electronics, textile and leather, infor­
from 95.19 to 99.54 MtCO2e, with the mean value of 97.54 MtCO2e, and
mation and communication, transport equipment, and etc. For example,
standard deviation of 0.54 MtCO2e. By 2040, these emissions range from
63 % of CO2 emission in 2017 of transport equipment sector originates
25.31 to 26.46 MtCO2e, with the mean value of 25.92 MtCO2e and
from imported products.
standard deviation of 0.17 MtCO2e. The uncertainty analysis of PBA and
It should be noted that the number of regions in this study is limited,
CBA emissions of the Italian electricity sector in 2017 and by 2040 is
including only two regions: Italy and RoW. While the number of regions
presented in Fig. 9.
has no impact on the PBA emissions because PBA is the direct emissions
The sensitivity analysis is conducted with updated IOT and NAMEA
from production and consumption activities of the national economy,
for 2019. The purpose of this analysis is to observe any change in the
the number of regions is expected to cause potential impacts on CBA,
model’s results due to economic interactions without altering the tech­
due to the changes in emissions from import and export. In case Italy
nology’s market share from 2017. A slight change is identified in the
imported goods and service from countries with higher emission in­
CO2 emission of the Italian electricity sector from CBA perspective,
tensity than the average emission intensity of RoW, the emissions
while there is no change in the PBA emission, as presented in Fig. 10.
embodied in import will increase, compared to the obtained results of
Specifically, the CBA CO2 emission of the Italian electricity sector re­
the existing model. Therefore, more detail of regions will bring a more
duces from 34.9 in 2017 to 34.3MtCO2e in 2019.
accurate obtained results on emissions embodied in trade activities and
Moreover, the national CO2 emission reduce in both PBA and CBA
CBA emissions.
perspectives as presented in Fig. 11. The PBA national emission in­
During the 2017–2040 period, the shares of air emissions embodied
creases by 1 MtCO2e from 2017 to 2019, which indicates the increase in
in import out of consumption increases, for example, from 40.8 % to
emissions of other economic sectors being connected to the electricity
49.5 % in CO2 emission. These increases occur in all particular economic
sector. In other words, if the electricity sector is not decarbonized, the
sectors. The largest CO2 emission increases are among electricity, in­
CO2 emissions will increase due to national production and consumption
formation and communication, and finance and insurance sectors, at
activities. In contrast, between 2017 and 2019, the CBA national emis­
around 11–12% points in the same period. This indicates that the trend
sion decreases by 4 MtCO2e, which originates from the trade activities
of transferring the national air emissions to other countries will continue
between Italy and RoW.
in the mid-term.
Five economic sectors’ consumption with relatively high shares of
embodied CO2 emissions are wholesale and retail, healthcare, food and

Fig. 8. Share of CO2 emission embodied in imports, relative to consumption, by year.

9
L.Q. Luu et al. Energy 290 (2024) 130109

Fig. 9. Uncertainty analysis of CO2 emissions of Italian electricity sector in 2017 and by 2040. Top left: PBA 2017, top right: CBA 2017, bottom left: PBA 2040,
bottom right: CBA 2040.

Fig. 10. Sensitivity analysis of CO2 emission of Italian electricity sector.

Fig. 11. Sensitivity analysis of CO2 emission of national economy.

10
L.Q. Luu et al. Energy 290 (2024) 130109

3.6. Limitations of the H-MRIO approach out the important contribution of imports in national air emissions.
Trade, pharmaceutical, computer and electronics, textile and leather,
Similar to IOA, H-MRIO is a static model. While time-differentiated information and communication, and transport equipment, are partic­
factors have been integrated into the model (power development sce­ ularly high importers of embodied carbon. This illustrates the limits of
narios, change in final demand, and a forecast of emission volume of the energy scenarios at the national scale, and supports the need for inter­
future years), there are some limitations belonging to the ‘static’ char­ national policy measures.
acteristics of the model. First, the model is based on fixed price, meaning The method and the practical application of the model show their
that there is no change in the price of products before and after the advantages in evaluating scenarios for the introduction and massive
integration of power development scenarios, as well as between 2017 deployment of RES in reducing GHG emissions in the energy sectors and
and 2040. Second, prices are the same for different power technologies. relevant economic sectors during the energy supply chain. However,
Though the purchase price of power is the same for all technologies, the there are scientific gaps which should be addressed in the near future,
production price should be different among electricity generation such as developing the method for integrating the changes in energy
technologies. efficiency, material efficiency and sector productivity into the quanti­
Moreover, the hybridization is a time-consuming process. For the fication process. Another area for future research is to extend the anal­
development of MRIO matrix, Istat and EXIOBASE data need to be ysis to various environmental impacts such as material consumption,
matched. Economic sectors of these two databases are classified differ­ and sustainability impacts, as mentioned in Ref. [44] which are very
ently; and there is also a variation in the level of sectoral aggregation. important in the current material shortage and sustainable development
For the hybridization of stressor matrices, the Ecoinvent and NAMEA context.
data need to be matched. Again, the different granularity between two
data sets requires time to match them, and in some cases, matching is Credit author statement
simply not possible directly. It is suggested that in future studies, the
level of aggregation should be more specified to reflect the diversity of Le Quyen Luu: Conceptualization, Methodology, Resources, Data
technologies in the electricity, gas and steam sector. curation, Writing – original draft, Writing – review & editing. Thomas
The integration of future power development scenarios utilizes the Gibon: Conceptualization, Methodology, Resources, Data curation,
similar and linear-forecasted intermediate flow matrix (for other eco­ Writing – original draft, Writing – review & editing, Supervision.
nomic sectors excluding the electricity sector). Though the future in­ Maurizio Cellura: Conceptualization, Resources, Writing – original
termediate flow matrix may be similar to the current one in the short draft, Writing – review & editing, Supervision. Eleonora Riva Sanse­
term, it is not convincing that in the long term, they will be similar. It is verino: Resources, Data curation, Writing – original draft, Writing –
expected that with the increase in energy efficiency, material efficiency review & editing, Supervision. Sonia Longo: Methodology, Resources,
and sector productivity, there will be a lot of (non-linear) changes in the Writing – original draft, Writing – review & editing. All authors equally
long term. Future studies should take into account these aspects when contributed to the paper’s development and writing. All authors have
forecasting the future intermediate flow matrix. read and agreed to the published version of the manuscript
Lastly, the study is restricted by data availability for power devel­
opment scenarios. Considering that recent EU policy such as the “Fit for Declaration of competing interest
55” package is internalized into national policy, in which the new
Emission Trading System that will be applied to private transportation The authors declare the following financial interests/personal re­
and buildings, there will be a change in the electricity sector, as well as lationships which may be considered as potential competing interests:
other energy-intensive industries such as transportation and buildings. Le Quyen Luu, Maurizio Cellura, Eleonora Riva Sanseverino reports
At the same time, the further requirement of climate targets will initiate financial support was provided by European Union. Thomas Gibon re­
more diverse low-carbon energy technologies to include battery energy ports a relationship with FNR project that includes: funding grants.
storage systems and hydrogen; as well as emission reduction options
related to social behaviour change. In that context, a comprehensive Data availability
(and updated) energy transition scenarios should be developed, and
extended from the electricity sector to economic-wide sectors. I have shared the link for the data and codes in the manuscript.

4. Conclusion and prospects Acknowledgements

In the context of climate risk with multiple proposals for decarbon­ Funder: This work has been partly developed in the framework of the
izing the energy sector, this paper introduced a H-MRIO model with project ’’Network 4 Energy Sustainable Transition –NEST″, code
integration of energy development scenarios for quantifying the sectoral PE0000021, CUP B73C22001280006, Spoke 1, funded under the Na­
and national air emissions. The method extends the quantification tional Recovery and Resilience Plan (NRRP), Mission 4, by the European
boundary beyond the energy sector to include other economic sectors Union – NextGenerationEU.
within the energy supply chain. Some of these economic sectors are The authors thank the SUSTAIN group at LIST for the fruitful
more affected by the decarbonization of the energy sector, while others methodological discussions. Thomas Gibon acknowledges funding from
bear fewer impacts. Additionally, the method highlights the relationship FNR project 12773213 on the assessment of future energy scenarios at a
between emissions embodied in trade activities, i.e. import and export, regional level.
and decarbonization of the energy sector.
The quantification results on the life cycle air emissions of the Italian
Appendix A. Supplementary data
electricity sector and national emissions from PBA and CBA perspectives
showed that during 2010–2040, the electricity sector’s air emissions
Supplementary data to this article can be found online at https://doi.
drop significantly due to changes in economic structure, emissions in­
org/10.1016/j.energy.2023.130109.
tensity and the decarbonization of the electricity sector itself. Moreover,
Italian air emissions are roughly halved, a trend that occurs in most
References
economic sectors. The forecasted energy transition not only shrinks
emissions in the electricity or energy sector, but reductions spread to [1] IRENA. World energy transitions Outlook 2023: 1.5◦ C pathway, 1. Abu Dhabi:
other economic sectors along the energy supply chain. The study points International Renewable Energy Agency; 2023.

11
L.Q. Luu et al. Energy 290 (2024) 130109

[2] IEA. The Critical Role of buildings. Perspectives for the clean energy transition. [23] Towa E, Zeller V, Achten WMJ. Circular economy scenario modelling using a
International Energy Agency; 2019. Accessed from, https://www.iea.org/report multiregional hybrid input-output model: the case of Belgium and its regions.
s/the-critical-role-of-buildings. Sustain Prod Consum 2021;27:889–904. https://doi.org/10.1016/j.
[3] European Commission. European Green Deal: commission proposes transformation spc.2021.02.012.
of EU economy and society to meet climate ambitions. 2021. [24] Gibon T, Wood R, Arvesen A, Bergesen JD, Suh S, Hertwich EG. A methodology for
[4] Climate Watch. Global historical emissions. 2022. integrated, multiregional life cycle assessment scenarios under large-scale
[5] IEA. Greenhouse gas emissions from energy: Overview. 2021. technological change. Environ Sci Technol 2015;49:11218–26. https://doi.org/
[6] IEA. Global energy review: CO2 emissions in 2021 14. 2021. 10.1021/acs.est.5b01558.
[7] Shindell D, Smith C. Climate and air-quality benefits of a realistic phase-out of [25] Li R, Zhang H, Wang H, Tu Q, Wang X. Integrated hybrid life cycle assessment and
fossil fuels. Nature 2019;573:408–11. https://doi.org/10.1038/s41586-019-1554- contribution analysis for CO2 emission and energy consumption of a concentrated
z. solar power plant in China. Energy 2019;174:310–22. https://doi.org/10.1016/j.
[8] Cellura M, Cusenza MA, Longo S. Energy-related GHG emissions balances: IPCC energy.2019.02.066.
versus LCA. Sci Total Environ 2018;628–629:1328–39. https://doi.org/10.1016/j. [26] Vélez-Henao J-A, Vivanco DF. Hybrid life cycle assessment of an onshore wind
scitotenv.2018.02.145. farm including direct and indirect services: a case study in Guajira, Colombia.
[9] Cellura M, Longo S, Mistretta M. The energy and environmental impacts of Italian J Environ Manage 2021;284:112058. https://doi.org/10.1016/j.
households consumptions: an input–output approach. Renew Sustain Energy Rev jenvman.2021.112058.
2011;15:3897–908. https://doi.org/10.1016/j.rser.2011.07.025. [27] Hertwich EG, Gibon T, Bouman EA, Arvesen A, Suh S, Heath GA, Bergesen JD,
[10] Cellura M, Di Gangi A, Longo S, Orioli A. An Italian input–output model for the Ramirez A, Vega MI, Shi L. Integrated life-cycle assessment of electricity-supply
assessment of energy and environmental benefits arising from retrofit actions of scenarios confirms global environmental benefit of low-carbon technologies. Proc
buildings. Energy Build 2013;62:97–106. https://doi.org/10.1016/j. Natl Acad Sci 2015;112:6277–82. https://doi.org/10.1073/pnas.1312753111.
enbuild.2013.02.056. [28] Wan L, Wang C, Cai W. Impacts on water consumption of power sector in major
[11] Kitzes J. An introduction to environmentally-extended input-output analysis. emitting economies under INDC and longer term mitigation scenarios: an input-
Resources 2013;2:489–503. https://doi.org/10.3390/resources2040489. output based hybrid approach. Appl Energy 2016;184:26–39. https://doi.org/
[12] Leontief W. Chapter 11: environmental Repercussions and the economic structure: 10.1016/j.apenergy.2016.10.013.
an input-output approach. In: Input-output economics. New York: Oxford [29] de Koning A. Creating global scenarios of environmental impacts with structural
University Press; 1970. 1986. economic models 201. 2018.
[13] Brandão M, Heijungs R, Cowie AL. On quantifying sources of uncertainty in the [30] Snam, Terna. Scenario national trend Italia. 2021.
carbon footprint of biofuels: crop/feedstock, LCA modelling approach, land-use [31] Istat. Dataset: IL Sistema DI Tavole input-output. 2021.
change, and GHG metrics. Biofuel Res. J. 2022;9:1608–16. https://doi.org/ [32] Eurostat. NACE Rev. 2 – Statistical classification of economic activites in the
10.18331/BRJ2022.9.2.2. European Community. 2008.
[14] Kennelly C, Berners-Lee M, Hewitt CN. Hybrid life-cycle assessment for robust, [33] Eurostat. Dataset: annual detailed enterprise statistics for industry. NACE Rev. 2, B-
best-practice carbon accounting. J Clean Prod 2019;208:35–43. https://doi.org/ E); 2022.
10.1016/j.jclepro.2018.09.231. [34] OECD. OECD economic outlook. 2022.
[15] Agez M, Muller E, Patouillard L, Södersten C-JH, Arvesen A, Margni M, Majeau- [35] Istat. Dataset: NAMEA air emissions, vol. 2. NACE Rev.; 2022.
Bettez G. Correcting remaining truncations in hybrid life cycle assessment database [36] Ecoinvent, n.d. Ecoinvent database Version 3.7. .
compilation. J Ind Ecol 2022;26:121–33. https://doi.org/10.1111/jiec.13132. [37] Eurostat. Manual for air emissions accounts, Eurostat manuals and guidelines.
[16] Suh S, Lenzen M, Treloar GJ, Hondo H, Horvath A, Huppes G, Jolliet O, Klann U, 2015.
Krewitt W, Moriguchi Y, Munksgaard J, Norris G. System boundary Selection in [38] Stadler K, Wood R, Bulavskaya T, Södersten C-J, Simas M, Schmidt S, Usubiaga A,
life-cycle inventories using hybrid approaches. Environ Sci Technol 2004;38: Acosta-Fernández J, Kuenen J, Bruckner M, Giljum S, Lutter S, Merciai S,
657–64. https://doi.org/10.1021/es0263745. Schmidt JH, Theurl MC, Plutzar C, Kastner T, Eisenmenger N, Erb K-H, de
[17] Leontief W. Chapter 7: multiregional input-output analysis. In: Input-output Koning A, Tukker A. Exiobase 3: developing a time Series of detailed
economics. New York: Oxford University Press; 1963. 1986. environmentally extended multi-regional input-output tables: EXIOBASE 3. J Ind
[18] Tukker A, Dietzenbacher E. Global multiregional input–output frameworks: an Ecol 2018;22:502–15. https://doi.org/10.1111/jiec.12715.
introduction and outlook. Econ Syst Res 2013;25:1–19. https://doi.org/10.1080/ [39] IEA. Energy statistics data browser. 2022.
09535314.2012.761179. [40] IEA. Greenhouse gas emissions from energy data explorer. 2022.
[19] Hong J, Zhang X, Shen Q, Zhang W, Feng Y. A multi-regional based hybrid method [41] Dietzenbacher E, Los B. Structural decomposition techniques: sense and sensitivity.
for assessing life cycle energy use of buildings: a case study. J Clean Prod 2017;148: Econ Syst Res 1998;10:307–24. https://doi.org/10.1080/09535319800000023.
760–72. https://doi.org/10.1016/j.jclepro.2017.02.063. [42] Cellura M, Longo S, Mistretta M. Application of the structural decomposition
[20] Teh SH, Wiedmann T, Castel A, de Burgh J. Hybrid life cycle assessment of analysis to assess the indirect energy consumption and air emission changes related
greenhouse gas emissions from cement, concrete and geopolymer concrete in to Italian households consumption. Renew Sustain Energy Rev 2012;16:1135–45.
Australia. J Clean Prod 2017;152:312–20. https://doi.org/10.1016/j. https://doi.org/10.1016/j.rser.2011.11.016.
jclepro.2017.03.122. [43] Roman-White SA, Littlefield JA, Fleury KG, Allen DT, Balcombe P, Konschnik KE,
[21] Wolfram P, Wiedmann T. Electrifying Australian transport: hybrid life cycle Ewing J, Ross GB, George F. LNG supply chains: a supplier-specific life-cycle
analysis of a transition to electric light-duty vehicles and renewable electricity. assessment for improved emission accounting. ACS Sustain Chem Eng 2021;9:
Appl Energy 2017;206:531–40. https://doi.org/10.1016/j.apenergy.2017.08.219. 10857–67. https://doi.org/10.1021/acssuschemeng.1c03307.
[22] Piñero P, Cazcarro I, Arto I, Mäenpää I, Juutinen A, Pongrácz E. Accounting for raw [44] Aghbashlo M, Hosseinzadeh-Bandbafha H, Shahbeik H, Tabatabaei M. The role of
material embodied in imports by multi-regional input-output modelling and life sustainability assessment tools in realizing bioenergy and bioproduct systems.
cycle assessment, using Finland as a study case. Ecol Econ 2018;152:40–50. Biofuel Res. J. 2022;9:1697–706. https://doi.org/10.18331/BRJ2022.9.3.5.
https://doi.org/10.1016/j.ecolecon.2018.02.021.

12