Transportation Research Part A 94 (2016) 514–531

Contents lists available at ScienceDirect

Transportation Research Part A

journal homepage: www.elsevier.com/locate/tra

Interplay between electricity and transport sectors – Integrating

the Swiss car fleet and electricity system
Ramachandran Kannan ⇑, Stefan Hirschberg
Laboratory for Energy Systems Analysis, Paul Scherrer Institut, 5232 Villigen PSI, Switzerland

a r t i c l e i n f o a b s t r a c t

Article history: Electric vehicles are seen as a future mobility option to respond to long term energy and
Received 18 August 2016 environmental problems. The 2050 Swiss energy strategy envisages 30–75% introduction
Received in revised form 9 October 2016 of electric cars by 2050, which is designed to support the goal of decarbonising the energy
Accepted 11 October 2016
sector. While the Swiss government has decided to phase out nuclear electricity, deploy-
Available online 1 November 2016
ment of electric cars can affect electricity supply and emission trajectories. Therefore,
potential interactions between the electricity and transport sectors must be considered
Keywords:
in assessing the future role of electric mobility. We analyse a set of scenarios using the
Electric-mobility
Electricity supply
Swiss TIMES energy system model with high temporal resolution. We generate insights
Integrated energy systems into cross-sectoral trade-offs between electricity supply and electrification/decarbonisa
TIMES tion of car fleets. E-mobility supports decarbonisation of car fleet even if electricity is sup-
Switzerland plied from large gas power plants or relatively low cost sources of imported electricity.
However, domestic renewable based electricity generation is expected to be too limited
to support e-mobility. Stringent abatement targets without centralised gas power plants
render e-mobility less attractive, with natural gas hybrids becoming cost effective. Thus
the cost effectiveness of electric mobility depends on policy decisions in the electricity sec-
tor. The substitution of fossil fuels with electricity in transport has the potential to reduce
revenues from fuel taxation. Therefore it is necessary to ensure consistency between elec-
tricity sector and transport energy policies.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Background

Electric vehicles are seen by many as a promising future mobility option that responds to today’s energy-economic-
environmental problems, such as increasing energy prices, climate change, inefficient resource usage, air and noise pollution
in urban areas, and so on (Høyer, 2008; THELMA; Arar, 2010; Srivastava et al., 2010; Delang and Cheng, 2012; Fulton et al.,
2009; Ernst et al., 2011; Brand et al., 2012; Dijk et al., 2013; Raslavičius et al., 2015; Seixas et al., 2015; Jenn et al., 2015).
Unlike alternative mobility options to address the aforementioned issues, e.g., hydrogen vehicles, which may require an
entirely new infrastructure, the advantage of electric mobility is in largely making use of the existing electric infrastructure
(Horst et al., 2009; Electrosuisse, 2015), although some upgrading and expansion may be necessary (Srivastava et al., 2010).
Another potential benefit of electric mobility is in the possibility of exploiting the electric storage batteries in electric vehi-
cles for managing the balance between electricity supply and demand. That is, electric mobility is also seen as a solution for

⇑ Corresponding author.
E-mail addresses: Kannan.ramachandran@psi.ch, r.kannan@email.com (R. Kannan).

http://dx.doi.org/10.1016/j.tra.2016.10.007
0965-8564/Ó 2016 Elsevier Ltd. All rights reserved.
 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531 515

Nomenclature

Aviation (D) domestic aviation

Aviation (I) international aviation
BAU business as usual scenario
BEV battery electric vehicle
CHF Swiss Franc
CHP combined heat and power generation
CO2 carbon dioxide
CROSSTEM cross-border Swiss TIMES electricity model
ESD energy service demand
ETS emission trading scheme
EU European Union
FC fuel cell
GDP gross domestic product
GTCC gas turbine combine cycle plant
HGV heavy goods vehicle
Hydro (D) dam storage hydro power
Hydro (P) pumped hydro power
Hydro (R)
run of river hydro power
ICE internal combustion engine
INT intermediate season
kW kilowatt
kWh kilowatt-hour
LC low carbon scenario
LGV light goods vehicle
MARKAL market allocation—modelling framework
PJ Peta Joule (1015 J)
PHEV plug-in hybrid electric vehicle
Rail (F) rail—freight transportation
Rail (P) rail—passenger transportation
RES reference energy system
Rp Rappen (cent)
SMR steam methane reformer
STEM Swiss TIMES energy system model
STEM-E Swiss TIMES electricity model
SUM summer season
TIMES The Integrated MARKAL EFOM System—modelling framework
t-km tonne kilometre
V2G vehicle to grid
vkm vehicle kilometre
WE weekends
WIN winter season
WK weekdays

supporting a high share of intermittent renewables (e.g., solar and wind); managing base load type power plant through off
peak charging, and to buffer the electric grid (i.e., vehicle to grid–V2G) for system balancing (Lund and Kempton, 2008; Horst
et al., 2009; Budischak et al., 2013; Borba et al., 2012).
The current Swiss energy system is highly dependent on imported transport (and heating) fuels (BFE, 2010), and is thus
incompatible with long-term climate change mitigation efforts. The 2050 Swiss energy strategy envisages 30–75% electric
cars by 2050 (Prognos, 2012). This is designed to support the goal of decarbonising1 the transport sector, given that the
car fleet alone accounts for 53% of transport sector energy demand (or 66% excluding international aviation and 18% of total
final energy consumption) (BFE, 2010). However, the well-to-wheel CO2 emissions (and primary energy use) depend on the pri-
mary sources of electricity supply. The current Swiss electricity is nearly decarbonised, with nuclear power contributing around
40%. However, the government has also decided to phase out this low-carbon source of electricity after the Fukushima Daiichi

1
Energy-related CO2 emissions in Switzerland were 44 million tons in 2010. About 40% of the CO2 were from the transport sector while the electricity sector
account for less than 10% (FOEN, 2012).
516 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531

nuclear accident (FASC, 2011; Leuthard, 2011), raising challenges for the deployment of electric cars without increasing
electricity-related CO2 emissions. In addition to the challenges for long-term generation, the deployment of electric vehicles
can affect power demand during periods of peak charging (van Vliet et al., 2012). Therefore, potential interactions between
the electricity and transport sectors must be considered in assessing the future role of electric mobility. Measures to reduce
CO2 emissions in transportation are often viewed as independent from the broader energy system (Lund and Münster,
2006). To our knowledge, the Swiss Energy Strategy (Prognos, 2012) adopted a similar sectoral approach, which may not fully
represent the important cross-sectoral trade-offs and synergies. For example, uptake and cost effectiveness of electric vehicle is
likely to be highly dependent on the cost of electricity supply, and the characteristics of the generation capacity. On the other
hand, configuration of electricity sector depends on electricity (kWh) and power (kW) demands. Understanding of the inter-
dependency and interplay between energy sectors requires a systemic approach. Analyses do exist with systemic approach
(Gül, 2008; Seixas et al., 2015). However, those analyses do not have adequate temporal resolution to capture the variability
of the electricity system.
Aim of this paper therefore is to show the interplay between the electricity and transport sectors and their trade-offs. A
comprehensive and flexible model of the Swiss energy system—the Swiss TIMES energy system model (STEM)—with high
intra-annual detail has been used (Kannan and Turton, 2014). A number of energy scenarios have been analysed with STEM
(Kannan, 2016). In this paper, we present the interplay between the electricity sector and private passenger car fleet under
different boundary conditions. Section 2 gives an overview of the STEM with key input data and assumptions. Definitions of
the scenarios and analytical results are presented in Section 3. The results are discussed from a policy perspective with key
conclusions in Section 4.

2. Analytical framework

The analytical framework used for the analysis is The Integrated MARKAL-EFOM System (TIMES) – a technology-rich, cost
optimisation modelling framework (Loulou et al., 2005). The model optimises (minimise) cost of technology and fuel mix to
meet a given energy service demands (ESD) based on competing energy pathways. In STEM, the full energy system of
Switzerland is depicted from resource supply to end-use ESDs. The model represents a broad suite of energy and emission
commodities, technologies and infrastructure. STEM has a modular structure for each end-use sector, energy resource sup-
ply, electricity generation, new and emerging fuel production options (e.g. hydrogen and biofuels) and infrastructure (fuel
distribution) (see Fig. 1). The model has a time horizon of 2010–2100 with an hourly representation of weekdays and week-
ends in three seasons (summer, winter, and an intermediate season). Thus the model has 144 intra-annual time slices. The
model is calibrated to 2010 Swiss energy statistics (BFE, 2010). As a single region model, spatial aspects of demands,
resources, infrastructure, etc. are aggregated. The overall structure of STEM and its input data/assumptions are well docu-
mented (Kannan and Turton, 2014). All cost details are declared in 2010 Swiss Franc2 (CHF2010) and we use a social3 discount
rate of 2.5% over the entire model time horizon to reflect the rate used in the 2050 Swiss Energy Strategy (Prognos, 2012).

2.1. Transport sector

The transport sector in the model covers the two broad transport service demand categories, viz. personal and freight
transport, which are quantified in terms of vehicle kilometre (vkm) and tonne kilometre (t-km). Fig. 1 shows a simplified
reference energy system (RES) of the transport module and its link to other modules. The transport model includes ten
modes of transport, though all the modes are not shown in Fig. 1. International aviation and military transport (others)
are not modelled in any detail, but are included for calibration to the Swiss final energy balance (BFE, 2010). To meet the
transport ESD, a wide range of existing and future vehicle technologies (e.g. cars, buses, and trucks) and fuel supply options
are depicted. A high level of detail is included particularly for the car fleet, with a wide range of alternative drivetrains and
fuels (see Table 1). The other transport modes (buses, trucks or rail) are depicted with a more limited number of alternative
technology and fuel options.
The transport module is calibrated for each mode of transport based on final energy use4 (BFE, 2010), annual vehicle kilo-
metres (Prognos, 2012), and fuel/energy efficiency5 (FOEN, 2012). The existing car fleet is aggregated into three fuel categories
viz. gasoline, diesel and natural gas.6 The aggregated fuel efficiencies for the existing car fleet are adopted from the Swiss
national greenhouse gas inventory (FOEN, 2012). All the existing cars are assumed to be retired linearly over the next 12 years,
i.e. all the existing cars are to be replaced after 15 years (BFS, 2012). In addition to the existing vehicle technologies, a range of

2
Indicative exchange rate: 1 CHF  1.04 US$ or 1.38 Euro (SNB—Swiss National Bank, 2014).
3
Consumers may apply a higher discount rate than society in selecting the optimal technology. Thus, we have sensitivity analyses with higher discount rates
(e.g. 10% in Kannan (2015)).
4
It should be noted that the transport fuel consumption in the Swiss energy statistics includes fuel tourism (BFE, 2013, 2010), i.e. cross-border tanking to
benefit from fuel price/tax differences. This fuel tourism is excluded for the estimation of ESDs based on greenhouse gas emissions inventory data (FOEN, 2012).
5
The fuel efficiency reflects the energy efficiency of vehicles, i.e. kilometre output per unit of fuel input. The fuel efficiency in the calibration year is the Swiss
national average efficiency that could also include efficient hybrid- and inefficient old cars.
6
The number of existing battery and electric plug-in cars is insignificant (<1%) (BFS, 2012) and therefore not modelled explicitly.
 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531 517

Other modules Fuel Vehicle Demands

distribution technology
Resource Personal
Car fleet transport
module
Diesel ICE

Gasoline Gasoline ICE

Freight
Gas ICE
transport
Refinery
Diesel Diesel Hybrid

Gasoline Hybrid
Natural Gas
Gas Hybrid
Fuel
conversion Hydrogen Hydrogen ICE Other end use
(e.g. Hydrogen, sectors
Hydrogen Fuelcell
biofules)
Plug-in hybrid
Biofuel
Battery Electric
Electricity
Electricity
supply
module Two
wheelers
Bus
LGV

CO2 Taxes
HGV

Rail

Aviation

Fig. 1. Simplified reference energy system of the transport module.

new and future vehicle technologies are represented with alternative fuel and drivetrain options. New vehicle technologies are
depicted in five-year vintages reflecting improvements in fuel efficiency and/or cost reductions. The technical characterization
and capital cost of cars are summarised in Table 1 (Densing et al., 2012). It is worth noting that car transportation is modelled as
a single demand, without distinguishing between different market segments (i.e. range and size7). In addition, each car tech-
nology is modelled as a representative car with similar performance characteristics (such that each type of drivetrain/fuel com-
bination represents an equivalent substitute in terms of performance). This means that STEM does not seek to model the choice
between a large and small car (since a cost optimisation framework is less suited for this purpose), but rather the choice of driv-
etrain or fuel. The model does not seek to represent switch from cars to public transport. However, demands for other modes of
personal transportations, such as bus, two wheelers, rail, etc. (and freight transportation) are explicitly modelled.
For the car fleet, two types of electric vehicles, namely, pure battery electric vehicles (BEVs) and plug-in hybrid electric
vehicles (PHEVs), are represented in STEM. For the PHEV, separate electric- and gasoline/diesel-mode efficiencies are imple-
mented. This enables us to choose pure gasoline mode, if the electricity cost is prohibitively expensive in a season or period.
For both types of electric cars (i.e. BEV, PHEV), the time of charging is unconstrained, but constraints are included to control
the rate of charging based on the existing infrastructure of 220 V and 16 A household fuses (i.e. about 3.5 kW). Moreover,
charging is only possible when the car is not being used (see Appendix C).

7
Changes to the size distribution of the car fleet over the model horizon can be specified as a scenario assumption, but this is not considered in this paper.
518 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531

Table 1
Characteristics of new car technologies. Source: Densing et al. (2012).

Car technology type# Fuel efficiency*** Capital cost

(km/GJ) (‘000 CHF2010 per car)
Fuel type Drive train 2010 2020* 2030* 2050* 2010 2020* 2030* 2050*
Gasoline ICE 332 2% 4% 4% 24 0% 1% 1%
ICE-AD 440 6% 13% 13% 24 2% 4% 4%
Hybrid 545 25% 67% 67% 29 3% 6% 8%
Diesel ICE 368 2% 4% 4% 26 0% 1% 1%
ICE-AD 474 8% 23% 23% 27 0% 0% 0%
Hybrid 575 26% 72% 72% 31 3% 6% 7%
Gas ICE 447 5% 11% 11% 25 1% 1% 1%
Hybrid 600 21% 54% 54% 30 3% 6% 8%
Electricity BEV 1409 5% 10% 10% 43 18% 28% 32%
**
Electricity/Gasoline PHEV 983 13% 30% 30% 35 8% 14% 17%
Hydrogen Fuel cell 1000 6% 13% 13% 42 8% 19% 29%
ICE 564 24% 63% 63% 32 3% 6% 7%

BEV–Full battery electric vehicle, HYB-hybrid vehicle, PHEV– Plug-in hybrid electric vehicles, ICE–Internal combustion engine, AD–advanced ICE.
*
Relative change from the vintage year 2010.
**
Combined efficiency based on gasoline (50%) and electric (50%) drive mode. STEM has the flexibility to use different share between these two modes,
but with a maximum of 85% of the annual distance covered by electric mode.
***
Fuel efficiency reflects the energy efficiency of car, i.e. vehicle kilometre output per unit of fuel input. For example, the fuel efficiency of 332 km/GJ for
gasoline ICE car implies an energy efficiency of 12.6 km per litre or 7.9 L per 100 km.
#
An annual driving distance of 14,000 km per year and lifetime of 12 years is assumed for all car technologies.

Similar to the car technologies, new and alternative vehicles are represented for other modes of transportation. Their
technical and cost data (see Appendix A) are adopted from analysis of global transport (Gül, 2008; Densing et al., 2012)
and other data sources (Kannan et al., 2007; ETSAP—The Energy Technology Systems Analysis Program, 2012).
Fig. 2 shows the assumptions on future transportation demands, which are consistent with the 2050 Swiss Energy Strat-
egy (Prognos, 2012). For aviation and ‘other’ (e.g. military applications) transport demands, kerosene and diesel demands are
directly adopted. It may be noted that the transport service demands are fixed (i.e. inelastic) and the model seeks to find the
least cost vehicle technology and fuel options.
For cars, an average diurnal demand pattern (i.e. time of driving) is estimated based on micro-census data (BFS, 2005) (see
Appendix C). We have normalized total annual car demand to follow this pattern and do not differentiate across the three
seasons. For the other transport modes only an annual demand is specified without a detailed demand curve at this stage.

2.2. Electricity supply

STEM represents all existing electricity generation capacity at an individual plant level (e.g. nuclear plants) or aggregated
by fuel and technology (e.g. run of river hydro, storage dam hydro). A range of new technologies (centralised gas power

Fig. 2. Relative change in transport service demand. Source: Prognos (2012).

 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531 519

plants, solar PV, geothermal, etc.) are included and their technical and cost data are documented (Kannan and Turton, 2014).
Electric interconnectors between the Swiss and the European electricity networks are modelled as flexible technologies so
that electricity can be imported and exported at any time. This approach enables the possibility of importing cheap off-peak
electricity, store the electricity via pumped storage and batteries in electric vehicles. Electricity stored via pumped hydro can
be exported during the times of higher international prices. However, electricity stored in vehicle is not fed back to grid, i.e.
vehicle to grid (V2G) option is not enabled at this stage. Electricity import (and export) prices are adopted from cross border
Swiss TIMES electricity model (Pattupara et al., 2014). We assume a set of electricity prices consistent with neighbouring
countries adopting a stringent climate policy, i.e. electricity is produced from low-carbon and renewable sources. Table 2
shows the range in electricity price assumptions and the hourly electricity price variation for the year 2050 is given in
Appendix B.

2.3. Other modules

A range of domestic and imported primary energy resources are represented in STEM. Imported fossil fuels include crude
oil, refined fuels and natural gas. The international energy prices from IEA’s Energy Technology Perspective (IEA, 2014b) have
been used. For fuels like diesel and gasoline, the price is estimated based on the historical (1970–2010) correlation between
international oil and refined fuel prices (BP, 2014; IEA, 2014a) (see Table 3). Domestic renewable resource potentials are also
implemented in STEM based on various sources and expert judgement (see Kannan and Turton, 2014). The model incorpo-
rates a range of technology options to produce hydrogen, e.g., from natural gas via steam methane reforming (SMR), electricity
via electrolysis, biomass/waste via gasification, among others. Similarly, a range of biofuel production options are depicted.
Fuels are supplied to end-use sectors via aggregated fuel distribution networks without any spatial details, given STEM is
a single-region model. An estimate of the existing stock of infrastructure is included based on the quantity of fuel delivered
in 2010; and is assumed to be retired linearly over the next 50–80 years. Then a range of taxes (e.g. CO2 tax, transport fuel
tax; electricity tax) are implemented (BFZ, 2013; EZV, 2014; Keller and Wüthrich, 2013). The European Union (EU) Emissions
Trading System (ETS) permit price of 14–51 CHF/t-CO2 for power sectors is assumed between 2010 and 2050 (Prognos,
2012). The current climate levy (1.5 Rp/l) and fuel tax (e.g., gasoline 0.87 CHF/l) are also applied for transport fuels
(Prognos, 2012; IEA, 2011; BFZ, 2013). For new and emerging transport fuel like natural gas, hydrogen, electricity, etc., an
energy tax similar to gasoline is implemented from 2020.

3. Scenario results and discussions

3.1. Scenario definitions

We analysed a number of energy scenarios (Kannan and Turton, 2014; Kannan, 2015, 2016). In this paper, we present two
core scenarios, business-as-usual (BAU) and low-carbon (LC60) and a few sensitivity analyses. The energy service demands
remain unchanged in all scenarios and sensitivities.

 The business-as-usual (BAU) scenario incorporates existing policies on the phase out of nuclear generation. Large-scale
centralised gas power plants are available. We assume an annual self-sufficiency in electricity supply i.e. no ‘net’ annual
import of electricity. This, however, does not restrict the timing of electricity exchange within a year.
 The low-carbon (LC60) scenario realises the emissions pathway of the New Energy Policy scenario of the Swiss energy
strategy (Prognos, 2012) – a 22% reduction in total (energy system wide) CO2 emissions by 2020 and 60% by 2050 (relative
to 2010, and including emissions from international aviation). It is important to note that no sector specific emission cap
is applied in the transport sector.

There remains some policy uncertainty in Switzerland over the future role of centralised natural gas power plants
(Prognos, 2012). Hence, we also explore cases without centralised gas power plants; this is denoted in the scenario name
with the suffix ‘NoCent’, (i.e. BAU-NoCent, LC60-NoCent). Sensitivities on international fuel price assumptions have been
tested using high and low energy price assumpitons corresponding to the IEA’s 6DS or 2DS scenarios, respectively (IEA,
2014b). Table 4 summaries the scenario definitions and names. The results from the core scenarios and sensitivities are dis-
cussed in the following subsections. It is to be noted that STEM is the whole energy system model and for this paper, the

Table 2
International hourly electricity price range.

Country 2020 2025 2030 2035 2040 2050

Hourly electricity price variations (Rp/kWh)
Austria 4–26 8–32 7–16 5–16 7–17 4–38
Germany 4–30 8–34 7–16 5–55 6–35 3–23
France 2–30 3–33 6–16 4–15 6–17 3–22
Italy 4–29 6–33 7–17 5–16 5–19 3–30
520 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531

Table 3
Fuel price assumptions. Source: IEA (2013, 2014b, 2014a), BP (2014) and own estimates.

Fuel type 2010 2015 2020 2025 2030 2035 2040 2050
CHF2010/GJ
Natural gas 7.9 10.3 11.8 11.9 12.2 12.6 13.2 13.8
Crude oil 14.6 18.0 18.4 18.9 19.7 20.9 21.7 22.8
Gasoline 16.0 20.3 20.7 21.3 22.1 23.4 24.3 25.3
Diesel 16.4 20.6 21.1 21.6 22.5 23.9 24.8 25.9

Table 4
List of scenarios and sensitivities.

Core scenarios ? Business as usual (BAU) Low carbon (LC60)

Sensitivities ;
No centralised natural gas power plants BAU-NoCent LC60-NoCent
High fuel price BAU-FP-H LC60-FP-H
Low fuel price BAU-FP-L LC60-FP-L

Fig. 3. Car fleet in the BAU scenario.

results and discussions are limited to the transport and electricity sectors. Results of other sectors are explained in the pro-
ject report (Kannan and Turton, 2014; Kannan, 2016).

3.2. Business as usual scenario

In the BAU scenario, total final energy demand declines at 0.9% per annum due to a combination of end-use energy effi-
ciency, fuel substitution/switching and uptake of building energy conservation measures. Though total final energy con-
sumption declines, the end-use electricity demand increases to 288 PJ by 2050 (from 215 PJ in 2010). The transport
energy demands decline about 44% (excluding international aviation) by 2050 from 2010 level. Of the 44% reduction in
the transport fuels, car fleet alone contributes to 37%. Fuel consumption in the car fleet is reduced to 40% by 2050 from
the 2010 level, despite increasing travel demand (see Fig. 2). Fig. 3 shows the car technologies in the BAU scenario. Existing
gasoline ICE cars are replaced by advanced ICE cars in the short term. Gasoline hybrid cars penetrate from 2020 and dom-
inate the rest of the modelling period. Since the oil price increases throughout the time horizon (Table 3), BEVs become
increasingly cost-effective towards 2050, by which time the cost of batteries is reduced by more than 30% (see Table 3).
In 2050, BEV accounts for about 40% of the car fleet (i.e. two million cars). The deployment of hybrids and BEVs results in
a 30% reduction in gasoline demand by 2035 and 60% by 2050. The average tailpipe (tank to wheel) CO2 emission of the
car fleet declines from 208 g-CO2/km in 2010 to 144 g-CO2/km by 2020 and 45 g-CO2/km by 2050 (black colour8 maker in
Fig. 3).

8
For interpretation of colour in Figs. 3–5, the reader is referred to the web version of this article
 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531 521

Fig. 4. Electricity supply in winter and summer seasons in the BAU scenario.

For other transport modes, improvements in efficiency offset some of the increases in demand leading to a moderate
reduction in fuel demand. Conventional ICE buses are replaced with hybrid buses in the short and medium term and hydro-
gen in the longer term. This hydrogen is produced mainly from natural gas. Heavy- and light goods vehicles begin to shift to
hybrid diesel engines from around 2030. However, it should be noted that the representation of these other modes is less
detailed than the car sector.

3.2.1. Electricity supply

Total electricity demand in the BAU scenario increases from 215 PJ in 2010 to 288 PJ in 2050, representing a 34% (or 0.7%
per annum) increase from 2010. The existing nuclear power plants are gradually replaced by natural gas turbine combined
cycle (GTCC) power plants and CHPs, to supply the increasing electricity demands. By 2035, one third of the electricity supply
is from gas plants. Hydroelectric generation remains roughly constant throughout the period, though additional pumped
hydro generation is deployed to profit from off-peak international electricity prices (see Appendix D). In the longer run,
renewables, like solar PV and wind, become cost effective because of reductions in capital cost of these technologies
(Kannan and Turton, 2012) and increasing natural gas prices (Table 3). By 2050, 12% of the electricity is generated from
non-hydro renewables in the BAU scenario.
A novel feature of STEM is its hourly time resolution, which provides additional insights into electricity schedule at hourly
level. Fig. 4 shows electricity schedule from the BAU scenario in 2050 on summer and winter weekdays and weekends.9 On
summer weekdays, electricity demand peaks to 8.4 GW at 6:00 and stays in the 7–8 GW range during the day until late evening.
The difference between the peak and lowest demand (which occurs at 2:00) is about 2.6 GW. The morning peak is due to charg-
ing of BEVs (brown shade in the export plot) using the imported electricity. On the supply side, base-load plants like run-of-river
hydro, gas, and CHPs contribute about 4.8 GW and the remaining demand is met with a combination of imported electricity
(orange shade), dam hydro, solar PV, and others. Since international electricity prices are relatively low on summer days
(see Appendix B), electricity imports are attractive from morning till noon. Some of this imported electricity is stored via
pumped hydro (shown with in light blue shade in the lower plot of Fig. 4) and BEVs. From 16:00, the stored electricity from
pumped hydro and other flexible sources of electricity generation (dam hydro and gas plants) are scheduled to supply the
demand; and the excess generation is exported. The summer weekday-exports eventually enable import during weekends
and in winter (to fulfil the self-sufficiency/no net import constraint). The import and export patterns and quantities are highly

9
In electricity schedule, electricity demand (blue line) and supply mix are shown in upper plot and the lower plots show electricity export (grey shade),
charging of BEVs (brown shade) and consumption by pumped hydro (zigzag light blue shade). The red line in the upper plots is the marginal cost of electricity
supply.
522 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531

dependent on the electricity price assumptions, which could also affect the choice of end-use technologies (e.g. the level of BEV
deployment). For example, when the imported-electricity is fully turned off, the electric cars are not deployed, instead the car
fleet continue to use gasoline hybrid cars (see Fig. 3).
On winter weekdays, electricity demand peaks at 11.4 GW at 6:00 again due to charging of BEVs. The demand pattern is
flatter than the summer weekday due to heating during daytime and nights. Nevertheless, the difference between peak and
lowest demand (1:00) is 2.9 GW. However, the overall electricity demand is far higher than in summer because of electri-
fication of some of the heating demand in the residential and services sectors. Compared to summer, the output from
run-of-river hydro plants is reduced but all base-load gas plants are scheduled at their full capacity. In addition, the contri-
bution from CHPs is relatively high due to high heating demands. The total output from base load plants (5.8 GW) is far
below the lowest demand, and flexible gas plants are operated to meet the remaining demand. During 1:00–8:00, a small
quantity of electricity (10% of the demand) is imported and used for charging BEVs. From 14:00, dam hydro is scheduled,
mainly for export.
On weekends, electricity demand is similar to weekday demand because of charging of BEVs. In both summer and winter,
cheap imported-electricity is used during weekends to meet the demands or stored via pumped hydro and BEVs. The elec-
tricity imports during the weekends facilitate export (of dam hydro outputs) during weekdays.

3.2.2. Carbon dioxide (CO2) emissions

Total CO2 emissions in 2050 reduce to 30 million tonnes (Mt-CO2) from 43 Mt-CO2 in 2010 (left panel of Fig. 5) – a reduc-
tion of 25% from 2010 level. In car fleet, direct (tailpipe) CO2 emission in 2010 is about 11 Mt-CO2, which reduces to 3 Mt-CO2
– a reduction of 70% (red bar in right panel of Fig. 5). In 2050, the e-mobility ‘shifts’ some of the CO2 emissions to the elec-
tricity sector due to gas based electricity generation (see Section 3.2.1). If the shifted-emissions to electricity sector were to
be allocated to car fleet based on the average emission factor of electricity supply, then the total emission from car fleet will
be 3.8 Mt-CO2 in 2050 (i.e. the sum of the tailpipe emission of 3 Mt-CO2 and emissions shifted to power sector of 0.8 Mt-CO2
(brown bar in right panel of Fig. 5). However, if the marginal CO2 emission factor from the gas plant is allocated to car fleet,
then shifted emission will be about 1.9 Mt-CO2 (green bar in Fig. 5). Thus, the total emission from the car fleet in 2050 will be
5 Mt-CO2 – a net CO2 reduction of 55% compared to 70% in tailpipe emissions.

3.3. Low carbon (LC60) scenario

In the LC60 scenario, final energy demand declines at an annual rate of 1.2% (vs. 0.9% in BAU). Electricity demands increase
to 287 PJ by 2050—a level similar to the BAU scenario. The transport sector’s fuel demand (excluding international aviation)
declines to 55% by 2050 vs. 44% in BAU scenario. The transport sector is highly electrified, particularly the car, bus and LGV
fleets, with electricity demand increasing to 56 PJ by 2050 (from 11 PJ in 2010). In the short and medium term, the technol-
ogy and fuel mix seen in LC60 scenario for the car fleet is similar to that in the BAU scenario (i.e., from ICEs to hybrid cars). In
the long term, the plug-in hybrid electric vehicles (PHEVs) penetrate first followed by the BEVs. By 2050, all cars are either
PHEV or BEV (see Fig. 6). The deployment of PHEVs results in earlier decarbonisation of the car fleet, with average emissions

Fig. 5. Sectorial CO2 emissions and CO2 from car fleet in BAU scenario.
 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531 523

Fig. 6. Car fleet in the LC60 scenario.

in 2035 declining to 70 g-CO2/km versus 84 g-CO2/km in the BAU scenario. By 2050, the car fleet is fully decarbonized on a
tank-to-wheel basis. The total energy demand of the car fleet in 2050 declines by 71% (vs. 60% in BAU) from the 2010 level.
In the LC60 scenario, the heavy goods vehicle (HGV) and light goods vehicle (LGV) fleets switch to hydrogen fuel by 2050
as the carbon constraint becomes very stringent. These developments lead to a concomitant decline in the consumption of
gasoline and diesel, contributing to a substantial reduction in CO2 emissions (Fig. 8).

3.3.1. Electricity supply

The electricity supply is similar to the BAU scenario in the medium term, i.e. gas generation replaces the retired nuclear
plants. As the carbon constraint becomes more stringent, renewable electricity generation becomes cost effective and
contributes 12% of the total supply by 2030; and 22% by 2050 (vs. 12% in BAU) (see Appendix D). The remaining demand
is supplied from gas-based generation since the domestic renewable potentials are fully exploited. The model chooses

Fig. 7. Electricity supply in winter and summer seasons in the LC60 scenario.
524 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531

Fig. 8. Sectorial CO2 emissions and CO2 from car fleet LC60 scenario.

base-load-type GTCC plants, which are more efficient than the flexible/dispatchable plants. The load variations are balanced
by electricity storage in BEVs and by adapting to operation patterns of dam hydro plants.
Fig. 7 shows the generation schedule in the LC60 scenario. On summer weekdays, electricity supply exceeds the demand
and the excess is exported. Dam hydro plants are mainly used for export during evenings and nights, while during 2:00–5:00
BEVs are charged with imported electricity.10 Compared to BAU scenario, (peak) the demand in the LC60 scenario is low due to
a lower load from air conditioning (because of the deployment of more efficient AC systems). The daytime peak is also curtailed
by the deployment of solar thermal systems for meeting hot water demand. However, a high peak demand still occurs in the
evening due to loads from AC and hot-water demand.
On winter weekdays, demand peaks in the mornings and evenings due to the large deployment of electric heat pumps for
space heating. Solar thermal systems supply a small quantity of heat during the day helping to reduce electricity demands
during 8:00–15:00. Thus, the LC60 scenario exhibits more predominant morning and evening peak demand compared to the
BAU scenario. CHPs significantly contribute to the winter demand as both electricity and heat demands are high. Again, a
large share of the output from dam hydro is used for the export market. Unlike in summer, BEVs are also charged during
the day time, which may be related to the availability of excess electricity from CHPs.

3.3.2. Carbon dioxide emissions

Fig. 8 shows the CO2 emission pathways in the LC60 scenario. The electrification of car fleet nearly decarbonises the car
fleet by 2050. However, the electricity sector accounts for half of the total emissions in 2050. The right hand panel of Fig. 8
shows the tailpipe CO2 emissions from car fleet and shifted CO2 based on the average and marginal emission factor of elec-
tricity supply. Though the tailpipe emission from car fleet is almost nil, the shifted-CO2 to the electricity sector is about 0.9–
3.9 Mt-CO2 in 2050 depending on emission allocation method. Thus, the net reduction from e-mobility is only 67–92%.

3.3.3. Cost of LC60 scenario

Compared to the BAU scenario, additional annual (undiscounted) costs in the LC60 scenario are about CHF 6.81 billion in
2050 (or 13% more than in the BAU scenario). Most of this additional cost occurs in end-use sectors. The additional cost in the
transport sector is about CHF 2 billion, which is mainly the vehicle costs. Given the reduced consumption of conventional
fuels in the LC60 scenario, system wide fuel costs and taxes decline by about CHF 2.4 and 1.3 billion respectively; and a large
share of this cost reduction occurs in the transport sector. Additional costs in the electricity sector are about CHF 2 billion
because of deployment of capital-intensive renewables. Total capital expenditure in the LC60 scenario alone increases to
about 8.7 billion compared to BAU. However, some of this additional expenditure is offset by reductions in fuel
expenditure/taxes.

10
On summer weekends, the BEVs are charged from solar PV outputs during the daytime and with imported electricity during evening and night. Again this
charging pattern is driven by assumptions on electricity import prices. The assumed international electricity price (in Table 2), and our implicit assumption that
there is an unlimited supply of electricity imports (or an unlimited market for exports) are highly uncertain and represent an area for further sensitivity
analysis.
 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531 525

Table 5
Selected indicators in 2050 from the scenarios and sensitivity analysis.

Indicators Units 2010 BAU BAU- BAU- BAU- BAU- BAU- LC60 LC60- LC60- LC60-
NoCent FP-H FP-L NoFuelTax NoImp NoCent FP-H FP-L
Final energy demand PJ 882 618 766 599 749 746 631 545 581 545 545
Share of transport energy in final % 32% 33% 28% 32% 30% 30% 34% 33% 37% 33% 33%
energy demand
Share of fossil fuel in final energy % 61% 28% 53% 23% 47% 50% 32% 13% 24% 13% 13%
demand
Energy demand for car fleet PJ 151 60 74 51 74 74 74 44 72 44 44
Total electricity demand PJ 215 288 220 307 241 235 269 287 229 287 287
Share of transport electricity in % 5% 12% 7% 16% 6% 6% 5% 19% 0% 19% 19%
total electricity demand
Total CO2 emissions M t-CO2 43 30 33 27 40 38 30 18 18 18 18
Share of transport CO2 in % 37% 23% 28% 18% 25% 26% 31% 10% 31% 10% 10%
total emissions
Share of power sector CO2 in % 2% 38% 0% 42% 26% 20% 30% 31% 5% 31% 31%
total emissions
Share of non-hydro renewable % 2% 13% 29% 18% 8% 12% 10% 22% 28% 22% 22%
in electricity supply
Average tailpipe CO2 emission g-CO2/vkm 208 45 81 19 81 81 81 2 59 2 2
of car fleet
CO2 emission reduction in % 0% 39% 27% 47% 23% 23% 27% 63% 45% 63% 63%
transport from 2010 level
Share of BEV % 0% 44% 0% 76% 0% 0% 0% 97% 0% 97% 97%
Cumulative (2015–2050) Trillion CHF – 2.26 2.27 2.30 2.17 1.98 2.28 2.37 2.42 2.40 2.32
undiscounted system cost

3.4. Results from the sensitivity analysis

Table 5 shows selected indicators from the scenario and sensitivity analysis from Table 4. Across the scenario, the trans-
port sector CO2 emission reduces between 23–63% compared to the emission reductions in the Swiss energy strategy SES
2050 of 40–60% by 2050. However, it should be noted that emissions in electricity sector have increased. In the BAU scenario,
about 30% of electricity is produced from natural gas based power plants (see Appendix D). If the GTCC plants are restricted,
distributed CHPs and non-hydro renewable electricity generation play a larger role (see BAU-NoCent in Appendix D). Given
finite domestic renewable resources and the restriction of net electricity imports (due to self-sufficiency), electricity
becomes a scarce commodity in the BAU-NoCent scenario. As a consequence, no electrification occurs in car fleet and the
gasoline hybrid vehicles dominate the market in 2050 (BAU-NoCent in Fig. 9).

Fig. 9. Comparison of car fleet in 2050 and average fleet level emissions.
526 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531

Fig. 10. Fuel demand and CO2 emissions from car fleet in 2050.

With high fuel prices (BAU-FP-H in Fig. 9) the share of BEVs increases to 70% by 2050 (vs. 40% in BAU) whereas BEVs are
unattractive in the low fuel price assumptions (BAU-FP-L).
The deployment of BEV in the BAU scenario is attractive in two ways: fuel tax and infrastructure expansion. The transport
fuel tax, which is more than half of the fuel cost, plays an important role in the choice of car technology. Since electricity for
transport is taxed based on energy, lower electricity demand per vkm due to high efficiency of BEV makes BEV a cost effec-
tive technology option. We performed a sensitivity analysis without fuel tax and the result reveals that BEV is unattractive
and the car fleet continues with gasoline hybrid vehicles similar to the low fuel price case.
Since electricity for BEV is produced from gas power plant it can be argued why gas cannot be used in cars. This is because
BEV avoids the need for expansion of gas distribution infrastructure. Even though new gas transmission infrastructure has to
be built for the new gas power plants the cost of expanding gas network for centralised electricity plants is lower than
expanding the gas distribution network to fuelling stations. In addition, the high efficiency of BEV means that the electricity
demand for car fleet is lower (than the equivalent fuel demand for gas car) and therefore the cost (if any) of expanding the
electric grid is also low. However, it may be recalled that STEM is a single-region model without any spatial representation
and T&D infrastructures are highly aggregated.
In the LC60 scenario, international energy prices have almost no impact on car fleet in terms of technology choice or fuel
mix, since the carbon constraint effectively determines the cost of using fossil fuels such as oil and gas.
Despite the carbon cap in the LC60 scenario, centralised gas power plants are deployed to facilitate decarbonisation of
end-use sectors (see Fig. 8). Restricting centralised gas power plants (LC60-NoCent scenario) leads to reduction in total elec-
tricity demand (Table 5) due to absence of alternative supplies, i.e. renewable potentials are assumed to be finite and net
imports of electricity are assumed to be unavailable. The electricity generation mix is similar to the BAU-NoCent scenario
with an increased contribution from decentralised CHPs using natural gas and woody biomass (see Appendix D). In the
absence of centralised electricity supply, the car fleet switches to natural gas hybrid vehicle (see Fig. 9) compared to BEV
in the LC60 scenario. The average tailpipe emissions decline only to 59 g-CO2/vkm in 2050 as against near zero in the
LC60 scenario. The other transport modes (buses and trucks) extensively switch to hydrogen fuel, with the hydrogen pro-
duced from natural gas.11
Across the scenarios, fuel demands in car fleet reduce to almost 50% by 2050 due to efficiency improvements in new cars
and fuel switching (Fig. 10). Tailpipe CO2 emissions from car fleet reduce from 11 Mt-CO2 in 2010 to between 0.1 and 5.4 Mt-
CO2 by 2050, a reduction of 50–98%. Due to e-mobility in some of the scenarios, CO2 emission associated with the car fleet is
shifted to electricity sector. If the emissions from electricity sector were to be allocated to car fleet based on average emission
factor, the net emission reduction from the car fleet is 50–92%. However, if the car fleet is assumed to be using the marginal
CO2 emission from the gas power plant, then the net maximum emission reduction is only about 51–67%. On the other hand,
most of the charging of BEV occurs during weekends and night time with imported electricity; and therefore it can be equally
argued that there is no hidden emission due to electrification of car fleet.

11
Centralized production of hydrogen from natural gas may be an attractive alternative for very ambitious mitigation targets. It is worth noting that the LC60-
NoCent scenario is an extreme scenario which requires the deployment of many exotic and expensive technology options (e.g., hydrogen in rail transport).
However, it is worth reiterating that some of the end-use sectors do not represent all the technology options available for some of the less significant demands.
 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531 527

4. Conclusions

This paper describes strong interactions between the transport and electricity sectors. The scenario analysis using the
Swiss TIMES energy system model identifies a number of key technology transitions in the long-term development of the
Swiss car fleet that are important for realising a range of energy policy goals. Across all scenarios, efficiency improvements
are seen through the deployment of gasoline hybrid vehicles. Hybrid vehicles represent a cost-effective technology choice in
the medium term across the scenarios analysed, which are likely to be realized with the help of continuing price signals
(along with incentives in the EU on vehicle standards).
The long-term transition of the car fleet depends however on the availability of relatively low cost source of electricity.
The centralised gas plants support the deployment of BEVs. If centralised gas-based electricity generation is not available,
then cars directly utilising natural gas are favoured. Thus, the cost effectiveness of electric mobility depends on policy deci-
sions in the electricity sector. Given that the car fleet accounts for 17% of the Swiss final energy use and 28% of the total CO2
emissions, future vehicle technology and fuel choice plays a crucial role in the development of the energy system. Across the
scenarios, CO2 emissions from the transport sector in 2050 reduce between 23–63% compared to the target of 40–60% in the
Swiss energy strategy. For comparison, the EU goal for mobility in 2030 is 40% greenhouse gas reduction (EU, 2014). In
Switzerland, e-mobility has the potential to decarbonise 30–93% of the car fleet and contributes to a reduction of 7–
10 Mt-CO2. However, the uptake of e-mobility faces a number of hurdles (Perdiguero and Jiménez, 2012), in particular
the availability of charging infrastructure (and also the availability of cheap electricity). This indicates possible role for policy
in supporting the initial development of charging infrastructure (Electrosuisse, 2012) and, where necessary, supporting grid
expansion.
Increasing electrification is resulting in continuous growth in electricity demands. Given the phase out of nuclear
generation, there is need for additional capacity in both the short and long term. Clear policy signals for electricity
sector are required to ensure this capacity is built to achieve low-carbon goal. This includes signals for continued
expansion of renewable generation. Moreover, it is essential to ensure consistency between electricity sector and
transport policies.
Realising the high deployment of some capital-intensive BEVs may be a daunting challenge given the high upfront capital
outlays faced by consumers. Though cost effective from social perspective, policy support may be necessary to provide
households and small enterprises with access to capital for investing in new efficient cars.
Substitution of fossil fuels with electricity and increasing efficiency of cars have the potential to lead to reduced revenues
from fuel taxation – a trend seen in Jenn et al. (2015). While this may be relatively insignificant over such a long timeframe, it
nonetheless implies a need to reduce expenditure or raise revenue from other sources.
International energy prices are also a key uncertainty affecting the future configuration of the car fleet. In BAU scenario,
low energy prices do not push e-mobility, nor require a major shift from conventional car. No e-mobility implies a lower
electricity demand and therefore less challenges to the electricity sector. However, such a scenario raises additional chal-
lenges to meet any climate change mitigation policy goals (and would increase dependence on imported fuels), and thus
is likely to demand additional policy intervention to support new technologies (gas, electric vehicles, etc.). On the other
hand, high energy prices induce more e-mobility (and indirectly support climate change mitigation). However, high energy
prices naturally imply higher energy system costs, which raise economic and social challenges. In either case, policy inter-
vention to lower barriers to the uptake of suitable technologies and support the conditions for investing in energy infrastruc-
ture are important.
The limited set of scenarios presented in this paper shed important insights into the development of the Swiss car
fleet; and its inter-dependency on the electricity sector. An outlook for further scenario analyses include: how depen-
dent is the future role of e-mobility on the availability of cheap electricity during night and weekends? This question
arises out of the results presented in this paper, which are derived from one set of (highly uncertain) assumptions on
electricity trade. Additional scenario analysis using different international boundary conditions on electricity trade can
be explored. Similarly, the current analysis did not consider driving range and charging time of BEV. We will further
develop the model to include sub markets within the car fleets (four different car sizes to reflect different drive
range).

Acknowledgements

The research reported in this paper was supported by the Swiss Federal Office of Energy and the SCCER-Mobility (Swiss
Competence Center for Energy Research - Efficient Technologies and Systems for Mobility). An earlier version of this paper
was presented at the IAEE European Energy Conference in Rome, October 2014.
528 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531

Appendix A. Characteristics of new vehicle technologies

Vehicle type Change in fuel efficiency (km/GJ) Change in vehicle costs (CHF/vkm)
2010 2020 2030 2040 2050 2010 2020 2030 2040 2050
HGV Diesel ICE 96 105% 110% 116% 121% 1721 96% 96% 96% 96%
HGV Hydrogen ICE 99 112% 125% 139% 155% 2621 82% 75% 71% 68%
HGV Hybrid diesel ICE 110 111% 120% 126% 133% 1838 95% 94% 93% 93%

LGV BEV 872 108% 113% 118% 123% 1913 96% 94% 89% 81%
LGV Diesel ICE 320 105% 110% 116% 121% 1140 100% 100% 100% 100%
LGV Hydrogen ICE 267 120% 143% 171% 205% 1698 81% 76% 73% 70%
LGV Gasoline ICE 236 105% 110% 116% 121% 1046 100% 100% 100% 100%
LGV Hydrogen FC 514 120% 139% 148% 164% 3701 44% 39% 37% 34%
LGV Hybrid diesel ICE 426 120% 137% 144% 151% 1252 98% 99% 98% 97%
LGV Hybrid gasoline ICE 315 120% 137% 144% 151% 1368 93% 92% 89% 88%
LGV PHEV diesel 830 109% 113% 118% 123% 1478 94% 93% 90% 89%
LGV PHEV gasoline 793 114% 118% 123% 129% 1296 92% 91% 88% 86%

Bus BEV 301 109% 113% 118% 123% 5105 91% 88% 85% 83%
Bus Diesel ICE 102 105% 110% 116% 121% 3676 98% 98% 98% 98%
Bus Hydrogen ICE 121 124% 154% 191% 238% 4204 90% 88% 87% 87%
Bus Hydrogen FC 168 124% 148% 158% 175% 6151 68% 66% 64% 63%
Bus Hybrid diesel ICE 169 124% 147% 154% 161% 4161 95% 90% 90% 89%

2-wheeler Battery EV 7504 108% 113% 118% 123% 1400 95% 94% 93% 91%
2-wheeler Gasoline ICE 765 121% 125% 125% 125% 1133 100% 100% 100% 100%
2-wheeler Hydrogen FC 4420 114% 129% 148% 164% 3221 47% 44% 41% 39%

Rail Diesel (passenger) 8 105% 117% 132% 151% 33,443 100% 96% 94% 94%
Rail Electric (passenger) 21 104% 109% 114% 120% 31,592 100% 100% 100% 100%
Rail Hydrogen FC (passenger) 13 106% 113% 121% 130% 41,759 82% 81% 80% 79%
Rail Diesel (freight) 7 103% 110% 118% 128% 17,176 98% 98% 98% 98%
Rail Electric (freight) 18 101% 102% 103% 105% 15,758 100% 100% 100% 100%
Rail Hydrogen FC (freight) 11 104% 107% 111% 115% 28,882 72% 65% 64% 62%
Source: Kannan et al. (2007), ETSAP—The Energy Technology Systems Analysis Program (2012) and Gül (2008).

Appendix B. Hourly variations in imported electricity price in year 2040 and 2050

The figures below show the variations in hourly electricity price assumptions for the year 2040 and 2050. They are from
CROSSTEM model (Pattupara et al., 2014) based on an assumption that the neighbouring countries adopt a stringent climate
policy, i.e., electricity is produced from low-carbon and renewable sources.
 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531 529

Appendix C. Hourly car driving pattern of Switzerland

Source: BFS (2005).

Appendix D. Comparison of electricity supply in 2050

Note to figure: Demand represents the end-user electricity demand excluding losses and electricity used in pumped
hydro plants. Electricity consumption of pumped storage is reported separately as ‘‘Pumps”. Output from pumped storage
hydro is 80% of its input. Gas (Base) and Gas (Flex) refer to base-load and flexible gas combined cycle plants respectively.
530 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531

References

Arar, J.I., 2010. New directions: the electric car and carbon emissions in the US. Atmos. Environ. 44 (5), 733–734. http://dx.doi.org/10.1016/j.
atmosenv.2009.09.042.
BFE, 2010. Schweizerische Gesamtenergiestatistik (Various Years 2000–2010). Bundesamt für Energie, Bern.
BFE, 2013. Analyse Des Schweizerischen Energieverbrauchs 2000–2012 Nach Verwendungszwecken. Bundesamt für Energie, Bern <http://www.bfe.admin.
ch/php/modules/publikationen/stream.php?extlang=de&name=de_6635532.pdf>.
BFS, 2005. Mobilität in Der Schweiz Ergebnisse Des Mikrozensus Zum Verkehrsverhalten. Bundesamt für Statistik <http://www.bfs.admin.ch/bfs/portal/de/
index.Document.90808.xls>.
BFS, 2012. Strassenfahrzeugbestand: Personenwagen Bei Treibstoff Und Jahr. Bundesamt für Statistik <http://www.pxweb.bfs.admin.ch/Dialog/statfile.asp?
lang=1&prod=11>.
BFZ, 2013. Tarif der CO2-Abgabe auf Brennstoffen: 60 Franken pro Tonne CO2 Verordnung über die Reduktion der CO2-Emissionen (CO2-Verordnung)
Änderung vom 7. November 2013.
Borba, Soares M.C.B., Szklo, A., Schaeffer, R., 2012. Plug-in hybrid electric vehicles as a way to maximize the integration of variable renewable energy in
power systems: the case of wind generation in northeastern Brazil. Energy 37 (1), 469–481. http://dx.doi.org/10.1016/j.energy.2011.11.008.
BP, 2014. BP Statistical Review of World Energy 2013 <http://www.bp.com/en/global/corporate/about-bp/energy-economics/statistical-review-of-world-
energy.html>.
Brand, C., Tran, M., Anable, J., 2012. The UK transport carbon model: an integrated life cycle approach to explore low carbon futures. Energy Policy 41, 107–
124.
Budischak, C., Sewell, D., Thomson, H., Mach, L., Veron, D.E., Kempton, W., 2013. Cost-minimized combinations of wind power, solar power and
electrochemical storage, powering the grid up to 99.9% of the time. J. Power Sources 225, 60–74. http://dx.doi.org/10.1016/j.jpowsour.2012.09.054.
Delang, C.O., Cheng, W.T., 2012. Consumers’ attitudes towards electric cars: a case study of Hong Kong. Transp. Res. Part D: Transp. Environ. 17 (6), 492–494.
http://dx.doi.org/10.1016/j.trd.2012.04.004.
Densing, M., Turton, H., Bäuml, G., 2012. Conditions for the successful deployment of electric vehicles—a global energy system perspective. Energy 47, 137–
149. http://dx.doi.org/10.1016/j.energy.2012.09.011.
Dijk, M., Orsato, Renato J., Kemp, René, 2013. The emergence of an electric mobility trajectory. Energy Policy 52, 135–145.
Electrosuisse, 2012. Getting Connected – Electromobility and Infrastructure, Fachgesellschaft e’mobile <http://www.e-mobile.ch/pdf/2012/
Getting_connected_GB_040912.pdf>.
Electrosuisse, 2015. 1000 Charging Stations in Switzerland <http://www.e-mobile.ch/index.php?pid=en,1,57>.
Ernst, C., Hackbarth, A., Madlener, R., Lunz, B., Sauer, D.U., Eckstein, L., 2011. Battery sizing for serial plug-in hybrid electric vehicles: a model-based
economic analysis for Germany. Energy Policy 39 (10), 5871–5882.
ETSAP—The Energy Technology Systems Analysis Program, 2012. E-TechDS—Energy Technology Data Source <http://www.iea-etsap.org/Energy_
Technologies/Energy_Technology.asp>.
EU, 2014. A Policy Framework for Climate and Energy in the Period from 2020 to 2030 <http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:
52014DC0015&from=EN>.
EZV, 2014. Belastung der Treib- und Brennstoffe (T 4.2) Stand on Feb 2014, Eidgenössische Zollverwaltung. <http://www.ezv.admin.ch/zollinfo_firmen/
04020/04256/04263/04528/index.html?lang=de>.
FASC, 2011. Federal Council Decides to Gradually Phase Out Nuclear Energy as Part of Its New Energy Strategy, Press Release. The Federal Authorities of the
Swiss Confederation, Bern. <http://www.admin.ch/aktuell/00089/index.html?lang=en&msg-id=39337> (25 May, 2011).
FOEN, 2012. Switzerland’s Greenhouse Gas Inventory 1990–2010 National Inventory Report 2011, Federal Office for the Environment, Climate Division,
3003 Bern, Switzerland. <http://www.bafu.admin.ch/climatereporting/00545/11894/index.html?lang=en>.
Fulton, D., Cazzola, Pierpaolo, Cuenot, François, 2009. IEA Mobility Model (MoMo) and its use in the ETP 2008. Energy Policy 37 (10), 3758–3768.
Gül, T., 2008. An Energy-Economic Scenario Analysis of Alternative Fuels for Transport. PhD Thesis Nr. 17888, ETH Zürich.
Horst, J., Frey, G., Leprich, U., 2009. Auswirkungen Von Elektroautos Auf Den Kraftwerkspark Und Die CO2-Emissionen in Deutschland. WWF Deutschland,
Frankfurt am Main <http://www.wwf.de/fileadmin/fm-wwf/Publikationen-PDF/wwf_elektroautos_studie_final.pdf>.
Høyer, K.G., 2008. The history of alternative fuels in transportation: the case of electric and hybrid cars. Util. Policy 16 (2), 63–71. http://dx.doi.org/10.1016/
j.jup.2007.11.001.
IEA, 2011. Energy Policies of IEA Countries: Switzerland. International Energy Agency, Paris.
IEA, 2013. World Energy Outlook. International Energy Agency, Paris.
IEA, 2014a. Energy Prices and Taxes <http://www.oecd-ilibrary.org/energy/energy-prices-and-taxes_16096835#>.
IEA, 2014b. Energy Technology Perspectives 2014: Harnessing Electricity’s Potential. International Energy Agency, Paris.
Jenn, A., Azevedo, I.L., Fischbeck, P., 2015. How will we fund our roads? A case of decreasing revenue from electric vehicles. Transp. Res. Part A: Policy Pract.
74, 136–147. http://dx.doi.org/10.1016/j.tra.2015.02.004.
Kannan, R., 2015. Aspirations for electrification: does the future electricity demand profile matter for electricity supply? – temporal aspects of energy
systems modelling. In: Hybrid Energy Modelling – Linkages and Interdisciplinarity, Annual wholeSEM Conference, 6–7 July, Cambridge, UK.
Kannan, R., 2016. Long term climate change mitigation goals under the nuclear phase out policy: the Swiss energy system transition. Energy Econ. 55, 211–
222. http://dx.doi.org/10.1016/j.eneco.2016.02.003.
Kannan, R., Turton, H., 2012. Cost of ad-hoc nuclear policy uncertainties in the evolution of the Swiss electricity system. Energy Policy 50, 391–409. http://
dx.doi.org/10.1016/j.enpol.2012.07.035.
Kannan, R., Turton, H., 2014. Swizerland Energy Transision Scenarios – Development and Application of Swiss TIMES Energy System Model (STEM). Final
project report to Federal Office of Energy, Paul Scherrer Institute, Villigen PSI <http://www.bfe.admin.ch/forschungewg/index.html?lang=de&dossier_
id=02886>.
Kannan, R., Strachan, N., Pye, S., Anandarajah, G., Balta-Ozkan, N., 2007. UK MARKAL Model Documentation <http://www.ucl.ac.uk/energy-models/models/
uk-markal/uk-markal-documentation>.
Keller, M., Wüthrich, P., 2013. Abschätzung Der Künftigen Entwicklung Von Treibstoffabsatz Und Mineralölsteuereinnahmen Grundlagenbericht. Infras,
Bern.
Leuthard, D., 2011. Sicherheit Hat Oberste Priorität, Presse- Und Informationsdienst UVEK. Bundeshaus Nord, Bern <http://www.uvek.admin.
ch/dokumentation/00474/00492/index.html?lang=de&msg-id=38101> (18 May, 2011).
Loulou, R., Remne, U., Kanudia, A., Lehtila, A., Goldstein, G., 2005. Documentation for the TIMES Model, Energy Technology Systems Analysis Programme.
<http://www.etsap.org/Docs/TIMESDoc-Intro.pdf>.
Lund, H., Kempton, W., 2008. Integration of renewable energy into the transport and electricity sectors through V2G. Energy Policy 36 (9), 3578–3587.
http://dx.doi.org/10.1016/j.enpol.2008.06.007.
Lund, H., Münster, E., 2006. Integrated transportation and energy sector CO2 emission control strategies. Transp. Policy 13 (5), 426–433. http://dx.doi.org/
10.1016/j.tranpol.2006.03.003.
Pattupara, R., Kannan, R., Turton, H., 2014. Alternative low-carbon electricity pathways in Switzerland and it’s neighbouring countries under a nuclear
phase-out scenario. Appl. Energy 172, 152–168. http://dx.doi.org/10.1016/j.apenergy.2016.03.084.
Perdiguero, J., Jiménez, J.L., 2012. Policy Options for the Promotion of Electric Vehicles: A Review. Research Institute of Applied Economics, University of
Barcelona <http://www.ub.edu/irea/working_papers/2012/201208.pdf>.
 R. Kannan, S. Hirschberg / Transportation Research Part A 94 (2016) 514–531 531

Prognos, 2012. Energieperspektiven für die Schweiz bis 2050. Energienachfrage und Elektrizitätsangebot in der Schweiz 2000–2050. <http://www.bfe.
admin.ch/php/modules/publikationen/stream.php?extlang=de&name=de_564869151.pdf>.
Raslavičius, L., Azzopardi, B., Keršys, A., Starevičius, M., Bazaras, Ž., Makaras, R., 2015. Electric vehicles challenges and opportunities: lithuanian review.
Renew. Sustain. Energy Rev. 42, 786–800. http://dx.doi.org/10.1016/j.rser.2014.10.076. Review Article.
Seixas, J., Simões, S., Dias, L., Kanudia, A., Fortes, P., Gargiulo, M., 2015. Assessing the cost-effectiveness of electric vehicles in European countries using
integrated modelling. Energy Policy 80, 165–176. http://dx.doi.org/10.1016/j.enpol.2015.01.032.
SNB—Swiss National Bank, 2014. Interest Rates and Exchange Rates November 2014 <http://www.snb.ch/en/iabout/stat/statpub/akziwe/stats/akziwe/
akziwe_S1_Wechsel>.
Srivastava, A.K., Annabathina, B., Kamalasadan, S., 2010. The challenges and policy options for integrating plug-in hybrid electric vehicle into the electric
grid. Electr. J. 23 (3), 83–91. http://dx.doi.org/10.1016/j.tej.2010.03.004.
THELMA – Technology-centered Electric Mobility Assessment <http://www.thelma-emobility.net/>.
van Vliet, O., Brouwer, A.S., Kuramochi, T., van den Broek, M., Faaij, A., 2012. Energy use, cost and CO2 emissions of electric cars. J. Power Sources 196 (4),
2298–2310. http://dx.doi.org/10.1016/j.jpowsour.2010.09.119.