The disruption effect of digitalization on the energy sector: a multimodal approach
1. WIR SCHAFFEN WISSEN – HEUTE FÜR MORGEN
The disruption effect of digitalization on the
energy sector: a multimodal approach
Stermieri Lidia:: PhD student :: Paul Scherrer Institut :: Energy Economics Group
WINTER 2020 SEMI-ANNUAL ETSAP MEETING
2. The digital society
Seite 2
E-learningE-book
Teleworking
E-commerce
E-banking
Smart buildings
E-retail
Smart energy
Digital practices
“Working from home can save energy and
reduce emissions. But how much?”**
“Residential energy consumption in
the United States has increased by an
estimated 6-8% compared with this
time last year”*
Rebound effect of the pandemic…
*The Covid-19 Crisis and Clean Energy Progress, IEA Report, June 2020
**IEA (2020), Working from home can save energy and reduce emissions. But how much?, IEA, Paris
1. Quantify the impact of digital practices on the energy
system
2. Analyze changes in user behavior
“Government lockdowns triggered a fall
of 50% to 75% in road traffic around the
world”**
How is society considered in the energy model?
3. The role of consumers’ behavior in the digital
transition: an agent-based model framework
Seite 3
Commuting
+ -
…
…
?
Commuting
Agent Social network
Decision process agent
Service sector
Decision process service Teleworking
?Impact on energy
demand
Increase digital
level sector
Transport
Residential
Service
4. Agent-Based Model
Seite 4
Households
Socio-demographicattributes
Decision process mechanism
New digital trends
Micro-level interactions
Households
Social interaction(network)
Macro-levelinteractions
Firms
Teleworking
E-learning
E-commerce
ABM
Exogenous demand
Technology share
Technology cost and attribute
Decision process:
• Believes, ideas , opinion, preferences
• Economic considerations (based on
income)
• Investment cost
• Operation cost
• Efficiency
• Energy cost
• Simulation on an annual basis
• Time horizon: 2020-2050
• Population growth over the time horizon
• Social network and interaction
Selection of the best technology
Households
Transport Residential ServiceSector
Agent
Aggregate
Technology
Annual mileage
Building type(MF,SF)
Heating/Electric consumption
Education
Job
Output
Energy demand
Technologies’ share
Variables
related to
teleworking
practice
Income
Age
Education
5. Integrated modeling approach
Seite 5
Agent-Based Model
Analysis of micro-level factors
that influence the diffusion and
adoption of a technology
• Social interactions
• Heterogeneity of
decision-making process
Swiss TIMES Energy System
Detailed representation of the
technological, economic and
environmental dimensions of
the energy system
Supporting long-term decision making and planning in the energy
sector considering individuals’ preferences and behavioral attitudes.
6. Swiss TIMES Energy Model (STEM)
Seite 6
Supply
Coal
Oil
Gas
Biomass
Primary
energydemand
Coal
Oil
Gas
Nuclear
Hydro
Bioenergy
Renewables
Imports&
Exports
Trade
matrices
Domestic
production
Transformation
Refinery
Gasprocessing
&distribution
Power
generation
Heat
production
Biomass
processing
Hydrogen
production
Syntheticfuels
Final
energydemand
Non-
energyuse
Industry
Transport
Residential
Services
Agriculture
Energy
servicedemand
Industrial
production
Industrialvalue
added
pkmtravelled
tkmtravelled
Househods&
householdsize
Valueaddedin
Services
CO2prices
Policies
Technologies
GDP
Population
Energyflows CO2emissions Investments
Leastcost
approach
STEM
Technologies costs and
attributes
Exogenous demand
Technology share
• STEM: Optimization model
• ABM: socio-economic model
ABM Coupling process with STEM
Aggregate technology:
• Non conventional
• Conventional
Transport Residential
Technology share Share constraints replace growth
constraints
ICE Gasoline
ICE Diesel
ICE Gas
Mild Hybrid Gasoline
Mild Hybrid Diesel
Hybrid Gasoline
Hybrid Diesel
Hybrid Gas
Plug-in Gasoline
Plug-in Diesel
Plug-in Gas
Battery electric
Hydrogen fuel cell
Electric boilers
Electric boilers with night storage
Heat pumps (electric)
Light fuel oil boilers
Gas boilers
Heat pumps (natural gas)
Coal process
Heat from CHPs on consumed on-
site
Heat from district heating
networks
Wood boilers
Pellet boilers
Solar boilers
Hydrogen boilers
Scenario analysis (horizon 2050)
Cost and attributes for aggregate technologies in ABM
Exogenous demand Dj
EXIT
YES
NO
START
∀𝐷𝑗 ∈ 𝑡𝑟𝑎𝑛𝑠𝑝𝑜𝑟𝑡, 𝑟𝑒𝑠𝑖𝑑𝑒𝑛𝑡𝑖𝑎𝑙, 𝑠𝑒𝑟𝑣𝑖𝑐𝑒 , ∀𝑡 ∈ 2030,2040,2050
𝐷𝑗,𝑡,𝑖 − 𝐷𝑗,𝑡,𝑖−1
𝐷𝑗,𝑡,𝑖−1
≤ ε
Dj energy demand
t milestone year
i current iteration number
Ɛ tollerance ( 0.05)
Iterative process
7. Case study
Seite 7
Baseline scenario (BAU) Climate target scenario (CLI)
8 Mt/CO2 in 2050
• Spread of practices in society
• Interdependencies between sectors
• Energy implication for long-term scenarios
• Society reaction to policy ( no awareness of
environmental issues by society)
1. Quantify the impact of digital practices
on the energy system
2. Analyze changes in users behavior
How is society reflected in energy model
assumption?
Assumptions for teleworking (If the agent adopts the practice of teleworking):
- 3 days/week of teleworking (commuting practice takes 24% of annual km in Switzerland)
- +20% in the agent’s heat consumption
- +10% in the agent’s electricity consumption
The effect of teleworking on energy demand and user behavior in Switzerland
8. BAU: insight from ABM
Seite 8
55.0
56.0
57.0
58.0
59.0
60.0
61.0
62.0
63.0
64.0
2020 2030 2040 2050
Total Bvkm (cars + public transport)
BAU ABM
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
2015 2018 2030 2040 2050
Share of Teleworking practice in Swiss
population
ABM Swiss statistics
Diffusion of teleworking:
• 43% of teleworking in 2050
8% Reduction in total Bvkm
Changes of transport mode due to
teleworking practice
0%
2%
4%
6%
8%
10%
12%
1st
iteration
2nd
iteration
3th
iteration
Absoluteerror
BAU convergence criteria
Bvkm 2030
Bvkm 2040
Bvkm 2050
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
2030 2040 2050
Bpkm share for public transport
BAU
ABM no teleworking
ABM with teleworking
13.20
13.40
13.60
13.80
14.00
14.20
14.40
14.60
14.80
15.00
15.20
15.40
2020 2030 2040 2050
Bvkm for working
BAU working Bvkm ABM working Bvkm
-5 Bvkm
-35% Bvkm for commuting
9. Coupled model: BAU scenario
Seite 9
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
BAU ABM BAU ABM BAU ABM
Share of aggregate technologies in private passenger
cars
Conventional Hybrid Ele
Hybrid vehicles not
selected in the ABM
Conventional
vehicles replaced
with electric vehicles
Increase of non
conventional
technologies in
residential sector
In the ABM, the adoption of heating technology
competes with the adoption of transportation
technology.
An agent adopting electric vehicle has less
disposable income to adopt a technology for
heating (natural gas boiler is the cheaper
technology)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
BAU ABM BAU ABM BAU ABM
Share of aggregate technologies in residential
demand
Conventional Non conventional
120
125
130
135
140
145
150
2020 2030 2040
Residential : Energy
demands (in PJ)
ABM BAU
+13%
2030
2030
2040
2040
2050
2050
10. Coupled model: BAU scenario
Seite 10
20
25
30
35
40
45
2015 2020 2030 2040 2050
CO2 emissions (Mt)
BAU ABM
+5%
- The reduction in transport and services sector does not compensate for the increase in
residential sector
- Conservative assumptions
- Need to consider interdependencies between sectors and rebound effects
651
133 128
213 232
151 149
0
100
200
300
400
500
600
700
BAU ABM
Final Energy Consumption in 2050 (PJ)
Industry Services Residential Transport
640
600
650
700
750
800
850
900
950
1000
2015 2020 2030 2040 2050
Primary Energy Consumption (PJ)
BAU ABM
+4%
Quantify the impact of digital practices on the energy system
11. CLI: 8 Mt/CO2 target 2050
Seite 11
0%
20%
40%
60%
80%
100%
2020 2030 2040 2050
ABM aggregate technology:
passenger cars(PJ)
Conv Non conventional
0%
20%
40%
60%
80%
100%
2020 2030 2040 2050
CLI aggregate technology:
passenger cars(PJ)
Conv Non conv
Different speed of technology adoption in transport!
Seite 11
-10 -5 0 5 10 15
CLI
ABM
Total CO2 emissions
Services Residential Transport
Industry CO2 captured in power generation CO2 captured in industry
CO2 captured in hydrogen CO2 captured in biogases/bioliquids Direct air capture
+20% conservation in Industry
0
20
40
60
80
100
120
140
160
CLI ABM
Transport total : Final consumption per
fuel in PJ
Hydrogen
Electricity
Biojet fuels
Biogas
Biodiesel
Ethanol
Natural gas
Jet fuel
Diesel
Gasoline
How to achieve 8 Mt/CO2 in 2050
+20% CO2
captured in
hydrogen and
biogases
12. CLI:CO2 tax and climate policy consideration
Seite 12
How much the carbon tax should be to achieve
CO2 reduction target and satisfy user
preferences?
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
2020 2030 2040 2050
CO2 emission Mt
CLI target ABM + CLI tax
To achieve 8 Mt/CO2 in 2050 without considering user
preferences:
2020 2030 2040 2050
CO2 tax CLI (CHF/Mt) 0.08 0.59 0.84 0.74
Considering user preference, the 8 Mt target is not achieved!
How is society represented in the energy model and in its assumptions?
13. • Open for discussion…
− CO2 tax must incorporate social behavior: how?
− The assumptions about the acceptance and evolution of society must be reflected in
the energy model:
− Are we overestimating future development?
− Are we underestimating cost?
Discussion
Seite 13
• Digital practices have a direct and indirect impact on different energy sectors
• Interdependencies between sectors must be considered to avoid
overestimation of impact
• Consumer preference and societal acceptance play an important role
14. Seite 14
Wir schaffen Wissen – heute für morgen
My thanks go to Dr. E. Panos and
Dr. T. Kober for their support in
my research
15. Digitalization and energy implication
*V. Court and S. Sorrell, “Digitalisation of goods: A systematic review of the determinants and magnitude of the impacts on energy consumption,” Environ. Res. Lett., vol. 15, no. 4, 2020