Content uploaded by Jan Remund
Author content
All content in this area was uploaded by Jan Remund on Jun 10, 2022
Content may be subject to copyright.
Department of the Environment,
Transport, Energy and Communication DETEC
Swiss Federal Office of Energy SFOE
Energy Research and Cleantech
Final report
FiPPS
Firm PV power generation for Switzerland
2/4
Date: 18 May 2022
Place: Bern
Publisher:
Swiss Federal Office of Energy SFOE
Research Programme Photovoltaics
CH-3003 Bern
www.bfe.admin.ch
Agent:
Meteotest AG
Fabrikstrasse 14
CH-3012 Bern
https://meteotest.ch
Authors:
Jan Remund, Meteotest AG, jan.remund@meteotest.ch
Michael Schmutz, Meteotest AG, michael.schmutz@meteotest.ch
Marc Perez, Clean Power Research, marcp@cleanpower.com
Richard Perez, State University of New York, rperez@albany.edu
SFOE head of domain: Stefan Oberholzer, stefan.oberholzer@bfe.admin.ch
SFOE programme manager: Stefan Oberholzer, stefan.oberholzer@bfe.admin.ch
SFOE contract number: SI/ 502286-01
The authors of this report bears the entire responsibility for the content and for the conclu-
sions drawn therefrom.
3/4
Summary
We investigate whether photovoltaics (PV) can effectively and economically contribute to a massively
renewable energy (RE) power generation future for Switzerland. Taking advantage of the country’s flex-
ible hydropower resources, we determine the optimum PV/battery configurations that can meet the
country’s growing electrical demand firmly 24x365 at the least possible cost while entirely phasing out
nuclear power generation. We examine several ultra-high RE scenarios where PV and hydro would
meet the bulk of the country’s demand. Depending on future cost predictions for PV and batteries, and
a small contribution from domestic or imported dispatchable resources, we show that power production
costs on the Swiss grid would range from 6 to 8 cents per kWh. This is well in line with market prices till
mid-2021 and strongly below the current price levels. Also, scenarios with no or only marginal imports
– either of electricity or e-fuels – would lead to only slightly higher costs – due to the effects of overbuild-
ing and curtailment. Our analyses show that firm PV power is an enabler of the energy transition and
can ease the energy trilemma – regarding security of supply, sustainability and affordability – existing
also in Switzerland.
Zusammenfassung
Wir untersuchen, ob die Photovoltaik (PV) einen effektiven und wirtschaftlichen Beitrag zur zukünftigen
Stromerzeugung aus erneuerbaren Energien (EE) in der Schweiz leisten kann. Unter Ausnutzung der
flexiblen Wasserkraftressourcen des Landes bestimmen wir die optimalen PV-/Batteriekonfigurationen,
die den wachsenden Strombedarf des Landes 24x365 zu den geringstmöglichen Kosten decken kön-
nen, während die Stromerzeugung aus Kernkraft vollständig eingestellt wird. Wir untersuchen mehrere
Szenarien mit extrem hohem EE-Anteil, bei denen PV und Wasserkraft den Großteil des Strombedarfs
des Landes decken würden. Abhängig von zukünftigen Kostenprognosen für PV und Batterien und
einem kleinen Beitrag von inländischen oder importierten regelbaren Ressourcen zeigen wir, dass die
Stromproduktionskosten im Schweizer Netz zwischen 6 und 8 Rappen pro kWh liegen würden. Dies
entspricht in etwa den Marktpreisen bis Mitte 2021 und liegt weit unter den gegenwärtigen Marktpreisen.
Auch Szenarien ohne oder mit nur geringfügigen Importen – entweder von Strom oder von E-Treibstof-
fen – würden nur zu geringfügig höheren Kosten führen – aufgrund der Auswirkungen von Überdimen-
sionierung und Abregelungen. Unsere Analysen zeigen, dass das Konzept der Überdimensionierung
und der Abregelung von PV-Anlagen die Energiewende möglich machen und das Energie-Trilemma der
Schweiz – bezüglich Versorgungssicherheit, Nachhaltigkeit und Bezahlbarkeit – entschärfen kann.
Résumée
Nous étudions si le photovoltaïque (PV) peut contribuer efficacement et économiquement à un avenir
de production d'énergie massivement renouvelable (RE) pour la Suisse. En tirant parti des ressources
hydroélectriques flexibles du pays, nous déterminons les configurations PV/batteries optimales qui peu-
vent répondre à la demande électrique croissante du pays, fermement 24x365, au moindre coût pos-
sible, tout en éliminant complètement la production d'énergie nucléaire. Nous examinons plusieurs scé-
narios d'ER très élevés dans lesquels le PV et l'hydroélectricité répondraient à la majeure partie de la
demande du pays. En fonction des prévisions de coûts futurs pour le PV et les batteries, et d'une petite
contribution des ressources dispatchables nationales ou importées, nous montrons que les coûts de
production d'électricité sur le réseau suisse seraient compris entre 6 et 8 centimes par kWh. Cela cor-
respond bien aux prix actuels du marché jusqu'à la mi-2021 et est fortement en dessous du niveau de
prix actuel. De même, les scénarios ne prévoyant aucune importation ou seulement des importations
marginales – que ce soit d'électricité ou de carburants électroniques – n'entraîneraient que des coûts
4/4
légèrement plus élevés – en raison des effets de la surconstruction et de la réduction des effectifs. Nos
analyses montrent que l'énergie photovoltaïque « ferme » est un facteur de transition énergétique et
qu'elle peut atténuer le trilemme énergétique (sécurité de l'approvisionnement, durabilité et prix abor-
dable) qui existe aussi en Suisse.
Main findings
- Overall, the results of the Energy Perspectives 2050+ could be confirmed. The optimum PV in-
stallation for this scenario (in this report #1) is 41 GW instead of the 37 GW modelled in the
perspectives including higher (14% instead of 9%) curtailment.
- The lowest costs result with about 40 GW PV, 15% curtailment and 15 GWh bat-teries, including
10% net imports (18 TWh during winter) a 10% rise of hydro power generation and storage (plus
1 TWh), a rise in pumped hydro (from 2.9 to 5.7 GW) and an import of 5 TWh of e-fuels (for
electricity generation).
- In all cases, required battery storage is low, amounting to 0.3 hours of full PV capacity in the case
of conservative cost assumptions, and ~1.2 hours in the case of optimistic cost assumptions.
- 10–85 GWh of batteries seem feasible compared to the expected electrical vehicle batteries,
which will include about 200 GWh of battery storage. Accessing 10% of this storage with by-direc-
tional loading systems would reduce the need of extra storage significantly
- As unlikely as this configuration may be, stand-alone grid operation would only increase these
costs by an average of 7% i.e., not constituting a showstopper.
- Without overbuilding and curtailment production costs would be an average of 63% higher across
all scenarios for the net-zero interconnected configuration, and 450% higher in the autonomous
grid configuration. The main factor for this cost difference is the amount of new battery storage
required that would respectively be 1300% and 7500% higher without PV oversize/curtailment.
- Overbuilding and curtailment of PV is an enabler: different levels of security of supply can be
reached without neglecting the net zero Co2 targets and still keeping electricity costs affordable.
The higher the level of security the higher the installed PV and the higher the share of curtailment
is needed.
- How to adopt the political and technical regulations to achieve the optimal values of overbuilt PV
is an open question and needs to be investigated.
Firm PV Power Switzerland
Making energy transition possible
2
Acknowledgements
This report was financed by the Swiss Federal Office of Energy (contract number:
SI/502286-01). The work is part of the activity 3.5 of the IEA PVPS Task 16.
Main contributions for this report were made by Marc Perez (Clean Power Re-
search USA), who modelled the results as well as Richard Perez (State Univer-
sity of New York at Albany), who gave general advice and wrote main parts of the
results chapter.
3
Contents
1 Introduction ................................................................................................. 6
2 Objectives .................................................................................................... 9
2.1 Key questions ...................................................................................... 9
2.2 Firm Power Generation ...................................................................... 10
2.3 Seasonal storage / flexibility ............................................................... 11
3 Definition of Swiss Energy system .......................................................... 12
3.1 Introduction ........................................................................................ 12
3.2 Today's system .................................................................................. 12
3.2.1 System data ................................................................................ 14
3.2.2 International grid connection ........................................................ 15
3.3 The situation in 2050 .......................................................................... 16
3.3.1 Cost levels 2050 .......................................................................... 17
3.3.2 Six scenarios ............................................................................... 19
3.3.3 Scenario #2 – 90% RES – 10% imported .................................... 21
3.3.4 Scenario #3 – 90% RES – 10% based on gas fired power plants –
limited imports ............................................................................ 22
3.3.5 Scenario #4 – 90% RES – 10% based on gas fired power plants
run by renewable H2 – limited import .......................................... 22
3.3.6 Scenario #5 – 84% RES – 10% imported and 6% based on gas
fired power plants ....................................................................... 22
3.3.7 Scenario #6 – 84% RES – 10% imported and 6% based on gas
fired power plants – with option of agri-PV .................................. 22
3.3.8 Overview of the scenarios ........................................................... 23
3.4 Order of redispatch ............................................................................ 24
3.5 Climate change .................................................................................. 25
4 Results ....................................................................................................... 26
4.1 General overview ............................................................................... 26
4.2 Implicit storage impact ....................................................................... 31
4.3 Sensitivity analysis: differences of meteorological years .................... 32
4.4 Conclusions ....................................................................................... 33
4.4.1 Costs of isolating Switzerland ...................................................... 33
4.4.2 E-Fuels ........................................................................................ 34
4.4.3 Policy and Market ........................................................................ 34
4.4.4 Comparison to other studies ........................................................ 35
4.5 Outlook .............................................................................................. 36
5 References ................................................................................................ 38
6 Annex ......................................................................................................... 40
4
Tables
Table 1: Hourly parameters of the period 2018–2020. ................................... 14
Table 2: Price assumptions for 2050. Round brackets: Lower costs based on
US studies. Square brackets: costs including share of agri-PV. For
PV and battery storage installation costs were used for modelling. For
non-optimized production types, the energy costs. ........................... 19
Table 3: Six scenarios used in this study. ...................................................... 19
Table 4: Annual electricity production in TWh 2018–2020 and 2050 with
scenarios 1–6. .................................................................................. 20
Table 5: Installed capacities in GW 2018–2020 and 2050 with scenarios 1–6.
Opt. means: optimized in this project. Seasonal Hydro storage
capacity is in TWh. ........................................................................... 21
Table 6: Overview of scenarios 1–6. .............................................................. 21
Table 7: Main results of the modelling for scenarios #1-#6 and #4a (no
import). ............................................................................................. 30
Table 8: Annual electricity imports in TWh with scenarios 1–6 with and without
isolation. ........................................................................................... 33
Table 9: Input definitions. ............................................................................... 40
Table 10: Average results of modelling based on 2018, 2019 and 2020
meteorological years. Imports will happen during winter time (imports
and exports are listed also in Table 9). ............................................. 41
Figures
Figure 1: Three phases of the energy transition towards wind and PV power. .. 6
Figure 2: The influence of PV overbuilding on firm power generation LCOE.
100% overbuilding means, that 50% of the theoretical PV production
is curtailed. ......................................................................................... 7
Figure 3: PV installed capacity (red line) and production (blue bars) 2018–
2020. ................................................................................................ 13
Figure 4: Modelled inflow to (seasonal) hydro storage expressed as potential
electricity power in MW. ................................................................... 15
Figure 5: Import and export of electricity to Switzerland 2018–2020 (positive:
import; negative: export). Black line: 15-days average. .................... 15
Figure 6: Supply-side electrical energy resources for all scenarios compared to
the current situation. The bottom part of the figure provides details for
the source labeled as ‘other’ in the top part. Note that scenario #4 is
the only scenario that does not include non-renewable (natural gas)
or possibly non-renewable (imports) resources. ............................... 23
Figure 7: Dispatch model applied in the Clean Power Transformation (CPT)
model. PSH stands for pumped hydro storage. ................................ 24
5
Figure 8: Annual dispatch of Swiss-based of supply-side resources for the year
2020. The top line of the stacked graph represents the Swiss grid
load .................................................................................................. 25
Figure 9: New PV capacities (top), optimal curtailment (middle) and new battery
storage for scenarios 1–6 for connected (left) and stand-alone
Switzerland (right). Max. acceptable shows the assumed maximum of
acceptable PV in Switzerland (55 GW). ............................................ 26
Figure 10: Swiss grid power generation costs for scenarios 1–6 and for
connected and stand-alone Switzerland. .......................................... 28
Figure 11: Annual dispatch of supply-side resources for the year 2020 illustrated
for the 100% renewable scenario with e-fuels (#4). The top graph
represents the net-zero interconnected configuration where winter
imports are energetically matched to summer export amounting to
net-zero. The bottom graph corresponds to the extreme stand-alone
grid configuration.............................................................................. 29
Figure 12: Share of energy production types for scenario #6 for 2050. ............. 30
Figure 13: Electricity production cost on the Swiss power grid as a function of PV
output curtailment for all scenarios. The top graph corresponds to the
interconnected grid configuration with net-zero import/exports with the
larger European grid. The bottom graph represents autonomous grid
configuration. Scenarios: #1: E-Perspectives 2050+; #2: 10% net
import; #3 and #4: 10% import, limited import capac., #5 and #6: 10%
import, 6% gas & agri-PV. ................................................................ 31
Figure 14: Comparing 2018, 2019 and 2020 electricity production cost on the
Swiss power grid as a function of PV output curtailment for scenario
#4. The top graph corresponds to the interconnected grid
configuration with net-zero import/exports with the larger European
grid. The bottom graph represents autonomous grid configuration. .. 32
6
1 Introduction
In a first and already historic phase of energy transition, wind and solar power
were expensive and rare. In the current second phase, wind and PV have be-
come cheap and normal. Production on country or transmission system level is in
most cases still lower than load. Curtailment happens, but most often only to omit
grid congestions. In the upcoming third phase, production will regularly be higher
than load. Production capacity, timing and storage becomes important and cur-
tailment will be a standard procedure.
Figure 1: Three phases of the energy transition towards wind and PV power.
As PV and wind production costs are currently and – according to almost all stud-
ies – will also in the future be lower than short- and long-term storage costs, the
main question is, which share of the energy should optimally be stored and which
share curtailed.
This is the question analyzed in this study. The results of this study will show the
amount of PV which is cost-optimally overbuilt (and curtailed) and stored in Swit-
zerland for a fully renewable energy system.
The calculations are based on the Swiss Energy Perspectives 2050+ scenario
net zero Basis (SFOE, 2021). The situation is modelled for 2050 only. This situa-
tion assumes a net PV production of 34 TWh, without nuclear production, but
with enhanced hydropower generation, as well as pumped hydro storage re-
sources for enhanced electricity load (electrified transport and building sector).
The Energy Perspectives 2050+ assume full (heat-pump) electrification of the
building sector’s heating requirements and large efficiency improvements (final
consumption for heating will be 37% lower).
The calculation is based on the analysis of 3 years (2018–2020). In addition to
exploring a fully Switzerland-based renewable scenario (PV, wind, hydro,
pumped hydro and other storage technologies), we will also explore scenarios in-
7
volving a degree of supply-side flexibility of up to 10–20% provided by e.g. dis-
patchable conventional generation (natural gas or e-fuels) and/or European grid
imports.
For each scenario analyzed, the results will consist of: Least-cost firm power lev-
elized cost of energy (LCOE), implied size of PV fleets, as well as implicit storage
(overbuilding) and real storage (beyond existing hydropower storage resource)
as defined in the Figure 2 below.
Figure 2: The influence of PV overbuilding on firm power generation LCOE.
100% overbuilding means, that 50% of the theoretical PV production
is curtailed.
While unconstrained PV is inexpensive (apparently below grid parity), firming PV
to meet demand 24/365 with storage alone (B) is unrealistically expensive. Over-
building of PV fleets reduces storage requirements to the point (C) (sweet spot)
where firm PV power generation can achieve true grid parity (D) (Perez et al.,
2021; O'Shaughnessy et al, 2021, Tong et al., 2021).
This work is focused on the guaranteed supply of electricity in every hour of a
year. It optimizes the costs in a macroeconomic viewpoint. There is no modelling
of grid (costs) and no market (merit order) modelling. We would like to mention
that to foresee the market model in 2050 is also highly uncertain. The current
marginal costs-based model is presumably not adequate for a system based on
marginal cost-free energy.
8
Results of previous investigations in the continental US (lower 48 states) and
tropical island power grids indicate that a 95% renewable energy based, opti-
mized wind/solar blend and an allowance for 5% supply-side flexibility via natural
gas could yield firm 24/365 LCOEs below 4 cents per kWh by 2040, with a
PV/wind overbuild of the order of 50% (Perez 2020; Perez et al., 2020, Tapaches
et al., 2020).
9
2 Objectives
The results show how much PV is cost optimally built in Switzerland to fulfil the
net zero 2050 goal of the Swiss Federal Council. This will help to define the
needed policy changes – push and pulls i.e. regulations and subventions – to
reach the targets regarding PV and storage. To know the amount of PV and stor-
age required is also important in order to find an optimal solution for future remu-
neration models – which will have to change from energy to power based. The
question of how to get enough incentives for PV and wind producers in case of
significant curtailment has to be answered as well.
Our power and not energy focused method shows the value of flexibility. The re-
sults will also show the cost of isolation of Switzerland from the neighboring
countries by calculating the option with limited or no electricity imports/exports to
neighboring countries (scenarios #3 and #4 and scenarios “a”).
Additionally, our results can show how much the option of building PV also on
farmland (and not only on rooftops) will change the optimum and the costs (sce-
narios #5 and #6). The results are important for grid operators as well as for pol-
icy makers and especially the government.
The main question to be answered is how to optimally overcome intrinsic intermit-
tency of PV.
2.1 Key questions
Optimal solutions are assessed in terms of:
• Optimum storage requirements — quantified in terms of installed PV ca-
pacity-hours.
• Optimum overbuilding — quantified as a percentage above unconstrained
PV capacity needed to meet energy requirements without curtailment.
• “Bottom line” LCOE of optimally configured PV — quantified in cents per
[firm] kWh.
• Different options to compare sensitivity of import and flexibility
We apply historical data from the Swiss transmission system operator (TSO)
Swissgrid (load data) and from the European association for the cooperation of
transmission system operators (TSOs) for electricity (ENTSO-E) (PV, hydro, nu-
clear, wind) as support to present and contrast the costs of achieving firm power
generation capable of entirely displacing existing conventional generation (nu-
clear energy in particular) and including also future electricity needs for transpor-
tation and heating.
10
We analyze firm forecasts and firm power generation from the standpoint of exist-
ing distributed PV fleets. Current installations are scaled up based on this spatial
distribution. The case study spans the years 2018–2020, for which we acquired
ENTSO-E historical hourly load data as well as PV, wind, hydro and nuclear pro-
duction as corrected for import and export by Swiss Energy Statistics from Swiss
Federal Office of Energy (SFOE).
2.2 Firm Power Generation
We calculate the real and implicit storage (achieved via overbuilding) require-
ments, as well as the corresponding capital cost premiums, and levelized energy
production costs (LCOE). In addition to the capital cost (CAPEX) of PV and stor-
age, LCOEs are also a function of the considered life cycle, the operation and
maintenance costs (OPEX) of PV and storage as well as the Weighted Average
Cost of Capital (WACC).
Real storage and implicit storage (overbuilding/curtailment) requirements are cal-
culated as a function of:
The capital costs of PV and storage.
• a future conservative scenario for 2050 with PV at CHF 860/kWp(stc)
(CHF 660/kWp for PV on farm land) and battery storage at CHF 330/kWh
(see chapter 3.3.1).
• An optimistic scenario based on US studies for bigger systems to model
cost sensitivity CHF 390/kWp and 45 CHF/kWh for battery storage)
Further assumptions:
• The round-trip efficiency of storage. We assume 90%.
• Since the objective is to supply the demand 24/7 at high-penetration,
there is no external battery recharge possibility at night or in off-hours.
Storage can only be recharged when renewable production exceeds de-
mand.
• We also consider flexibility defined in terms of the fraction of energy al-
lowed from external, non-renewable sources and/or import. This external
source could be supply-side, e.g., from legacy or new natural gas units,
and/or demand-side from load management. We consider flexibility levels
of 0% to 10%. The seasonal storage via e-fuels (H2, methane) will be
modelled via this flexibility (see below).
The financial specifics assume:
• A 30-year life cycle;
• Operation and maintenance costs of 1% of CapEx per year for PV;
11
• Operation and maintenance costs of 0.1% per full cycle for battery stor-
age
• A 3% Weighted Average Cost of Capital, representative of the utility in-
dustry.
For a given time horizon, location, and PV fleet configuration, the cost of firm PV
power generation is obtained by extracting the lowest life-cycle cost combination
of storage and overbuilding, sufficient to meet the firm forecast requirements.
We calculate storage and implicit storage requirements to firmly supply the de-
mand of Switzerland in its entirety. We apply the Clean Power Research Clean
Power Transformation (CPT) model (Perez et al. 2019) to derive the optimum
combination of real and implicit storage leading to the lowest possible firm gener-
ation cost.
2.3 Seasonal storage / flexibility
The implicit storage approach underlying the proposed study seeks the optimum
solution between storage and over-build/curtailment given an allowable amount
of flexibility (Perez et al., 2020). The required size and duration of storage are a
direct result of this optimization. The ongoing investigations in the US show that
the duration of storage (hence the need for seasonal reserves) is greatly influ-
enced by both oversizing and allowed flexibility. In particular, supply-side flexibil-
ity (import/export, gas fired power plants with natural gas and renewable energy-
based hydrogen or methane) alleviate the needs to build up long-term reserves
for extreme low supply/high demand situations. Optimum solutions show that
seasonal-duration storage is not needed to supply competitively priced firm
power and meet demand year around.
The six simulation options outlined in chapter 3 will capture optimum require-
ments for Switzerland and characterize any long-term storage requirements if
needed.
For hydro power we assume a growth of the seasonal storage of 2 TWh for sce-
nario #1 and 1 TWh for scenarios #2–#6 (Table 3).
12
3 Definition of Swiss Energy system
3.1 Introduction
The Swiss Energy System is defined in Table 9 in the Annex. Here we give an
explanation of the terms and values used.
The existing system is based on the Swiss electricity statistics and hourly data of
ENTSO-E1 between 2018 and 2020. The numbers can be used to be scaled up
for future scenarios. As the ENTSO-E source includes some missing values, it
has been corrected to the Swiss electricity statistics2. PV production had to be
gap filled as well. This was done with the aid of Swissmetnet stations, averaged
and modelled to a 15° tilted plane.
The future system (2050) is based on the Swiss Energy Perspectives 2050+
(SFOE, 2021). This includes several scenarios of possible future energy systems
fulfilling the climate agreement of Paris (1.5°C target).
3.2 Today's system
Today's system is defined as the average of the years 2018–2020.
Yearly production in TWh is given as well as installed capacities and cost levels
(in cts/kWh). Gross production is 70 TWh, net production 66 TWh. The losses are
based on consumption of pumps for hydro power and on grid losses.
The system is defined by a high share of hydro power. This is separated into
three types:
1. Hydro storage (large dams in the Swiss Alps mainly for seasonal storage),
2. hydro pumped storage (mid-sized dams often combined with large seasonal
storage dams) to store energy for some hours or days and
3. run of river hydropower system (of the rivers flowing from the Alps to the bor-
ders).
New renewables are relatively small. PV is at 2.2 TWh, wind production at 0.15
TWh. PV installations are growing at a rate of about 30% annually; annual incre-
ment of installed PV needs to be enhanced by a factor of three (from 0.5 to at
least 1.2 GW/year) to achieve the goals of net zero policy.
1 Source: https://transparency.entsoe.eu/
2 https://www.bfe.admin.ch/bfe/de/home/versorgung/statistik-und-geodaten/energiestatisti-
ken/elektrizitaetsstatistik.html
13
Figure 3: PV installed capacity (red line) and production (blue bars) 2018–
2020.
Nuclear production is 24 TWh. Four nuclear power stations are running. The con-
struction of new nuclear is forbidden by law. There is no fixed deadline for the
phase out of the nuclear power stations. However, as they are built between
1969 and 1984, they are already relatively old. Life time is expected to be further
extended and foreseen between 50 and 60 years. In 2050 the scenarios see no
nuclear power stations and production.
Currently, electricity has a share about 25% of the energy consumption in Swit-
zerland. 75% are non-renewables – all imported. In future (2050) this will change.
The main scenario is highly based on electricity.
Since many years Switzerland is exporting electricity on the annual level. Those
exports happen during summer time. In winter Switzerland is importing electricity.
This imbalance will grow when nuclear power is replaced mainly with PV.
Swiss electricity production does not follow the Swiss load. Swiss hydro power
plants (storage and pumped storage) are still exporting a certain amount of elec-
tricity to the surrounding countries during peak hours (morning and evening).
14
3.2.1 System data
The current system is defined by hourly values of three years 2018–2020 (Table
1) and includes the following parameters.
Table 1: Hourly parameters of the period 2018–2020.
Parameter Abr. Source Remark
Load L ENTSO-E Actual generation per production type
Nuclear PN ENTSO-E Actual generation per production type
Pumped hydro -
storage
PHp ENTSO-E Actual generation per production type
Hydro storage
(dams for sea-
sonal storage)
PHs ENTSO-E Actual generation per production type
Hydro run of
river
PHr ENTSO-E Actual generation per production type
Wind PW ENTSO-E Actual generation per production type
PV PPV ENTSO-E
Swissmet-
net
Required a strong correction as in 2018 only
a few PV installations were covered – and
the coverage rose significantly till 2020
Gap filled with average of 20 Swissmetnet
station data modelled to 15°S inclination
PV installed ca-
pacity
CPV SFOE Modelled to hourly data
Import PI ENTSO-E Cross-border physical flow between Switzer-
land and the neighboring countries
Rest R Modelled R = L - PN – PHp – PHs – PHr – PW – PPV - PI
Pumped hydro –
consumption
LHp
Modelled Negative part of rest (< -50 MW ); sum of
pump load & consumption < 2900 MW and
scaled to match annual consumption
Hydro storage
filling state
CHs SFOE Modelled from weekly to daily state
Inflow to hydro
storage (net)
PHsiN Modelled Delta of filling state. Shows approximately in-
flow due to snow melt
Inflow to hydro
storage (gross)
PHsiG Modelled Delta of filling state plus PHs, smoothed over
24 hours and scaled up to match yearly PHs
production
15
Figure 4 shows the modelled inflow data – net and gross.
Figure 4: Modelled inflow to (seasonal) hydro storage expressed as potential
electricity power in MW.
3.2.2 International grid connection
The European grid was started in 1958 on the Swiss border in Laufenburg. There
the first lines were built between Germany, Switzerland and France. The Swiss
grid is still highly interconnected with neighboring countries3. The electricity flow-
ing through Switzerland is in the range of 50% of the electricity consumption
within Switzerland. Italy depends heavily on the flow mainly from Germany. Elec-
tricity is generally imported in Switzerland during winter half year and exported
during summer (up to 4.5 GW import and 8 GW export).
Figure 5: Import and export of electricity to Switzerland 2018–2020 (positive:
import; negative: export). Black line: 15-days average.
Switzerland currently has no bilateral agreement with the EU regarding electricity
due to a missing institutional agreement. Therefore, the market integration is lim-
3 https://www.entsoe.eu/data/map/
16
ited and the outlook uncertain. Swiss utilities still take part in the day ahead mar-
ket EEX4. However, a part of the electricity balancing market and renewable cer-
tificates market are not open for Swiss companies.
Historically, the electricity price is defined during summer by the German / French
market (EEX) and during winter by the Italian market. End customers pay about
15–25 cts/kWh. Market price during the last 10 years was about 5 cts/kWh. How-
ever, during winter 2021/22 the prices for day ahead electricity rose up to 20
cts/kWh.
The market price is based on old, amortized power plants. During the last 20
years this price was too low to make investments for new power plants economi-
cally interesting (called missing money problem of the European market system).
The near future is relatively uncertain. The new regulation by the EU, to reserve
70% of the capacity to cross-zonal electricity trade, poses new challenges to
Switzerland5.
For all six scenarios (defined in chapter 3.3) we also modelled the extreme condi-
tions of isolating Switzerland. Those scenarios are named with an “a” (e.g. 1a).
3.3 The situation in 2050
In 2021, the Swiss government published an update of the Energy Perspectives
– called 2050+6 (SFOE, 2021). This report shows possible pathways to a climate
neutral energy system. In this study we use the main scenario "ZERO Basis".
New (non-hydro) renewables will grow from 3 to 40 TWh. PV has and shows the
biggest potential with 33 TWh. According to the Swiss Federal Office of Energy
(SFOE) and based on their solar cadaster, 67 TWh of electricity can be produced
on buildings. In reality, the rooftop potential is presumably lower and in the range
of 50 TWh / 50 GW7. In this study, a cap of 55 GW is applied which includes
about 40 GW for rooftop and 15 GW of installations aside buildings (e.g. agri-PV,
parking sites, floating PV). Wind energy potential would lie in the range of 8 TWh.
However due to high population density this shrinks to 4.3 TWh, which even
seems rather at the upper end of the possible contribution. Therefore, we used
50% lower values for wind power for scenarios #2–#6.
Also, hydro power is foreseen to grow. Mainly seasonal storage and hydro
pumped storage capacities would be added. Seasonal storage is enhanced from
10 to 12 TWh (scenario #1) or 11 TWh (scenarios #2–#6) according to the official
4 https://www.eex.com/en/market-data
5 https://www.swissgrid.ch/en/home/operation/market/european-market.html
6 https://www.bfe.admin.ch/bfe/en/home/policy/energy-perspectives-2050-plus.html
7 https://magazin.nzz.ch/nzz-am-sonntag/wirtschaft/solarenergie-ehrenrettung-
ld.1679852?mktcid=smsh&mktcval=Twitter
17
targets and the Round Table discussions8. The price for these new systems is
relatively high and would need special investments / securities by the govern-
ment. As the scenarios show differences regarding to amount of new hydro the
costs are also slightly varied.
PV off buildings is only included in one of the scenarios and discussions about
feasibility just started in Switzerland. Main reasons for this are high population
density, high price of land and the high importance of landscape protection in
Switzerland. Most presumably Federal Act on Spatial Planning (RPG) and sub-
vention regulations for agriculture need to be adopted. In scenario #6 a part of
PV (around 30% or 13 TWh) is produced on farm land with agri-PV installations.
The official naming in the Energy Perspectives 2050+ is that 100% of the energy
is produced within Switzerland. However, this isn't fully correct. The report in-
cludes 13.6 TWh of imported liquids based on Power to X technologies (PtL,
based on renewables). Therefore, the share of energy produced in Switzerland is
84%. Additionally, the fuels for aircrafts are not included. About 20 TWh of re-
newable PtL is used for air transport at levels of 2019. Keeping the same levels
of air transport, the real share of energy produced in Switzerland is 72%. None-
theless, we use the term 100% in this report not considering the imported PtL
and aside usage for air transport.
According the scenario ZERO Basis a small part of hydrogen is produced within
Switzerland (1.9 TWh; to produce this 7.4 TWh of electricity is needed). The sce-
nario is rather optimistic regarding efficiency measures. Total energy consump-
tion will not grow much. One of the realistic reasons is that today many electric
heating systems exist, which will be exchanged by heat pumps (saving 70% of
the electricity). Oversizing was included in the modelling to a limited extent. In the
scenario, 37 GW of PV is foreseen with 33.6 TWh of production. As 37 GW in
Switzerland produce on average 37 TWh of energy peak shaving of 9% is in-
cluded.
3.3.1 Cost levels 2050
There are no cost assumptions per technology published in the Energy Perspec-
tives reports. The report only includes some general macro economical figures.
The definition of price levels 30 years ahead includes high uncertainties. Four dif-
ferent sources have been used as a basis: published papers (Figgener et al.,
2019, NREL ATB11), Nexus-e reports (ESC, 20209), conferences (EES 202110)
and selected Swiss experts, which have been interviewed. The reported values
8 https://www.admin.ch/gov/de/start/dokumentation/medienmitteilungen.msg-id-86432.html
9 https://nexus-e.org/documentation/
10 https://www.ees-europe.com/
18
were included in the definition. As all PV (aside scenario #6) includes only rooftop
PV and labor costs in Switzerland are high, the LCOE of PV will stay rather high.
The most comprehensive work on costs and cost perspectives exists in the An-
nual Technology Baseline (ATB) of the US National Renewable Energy Labora-
tory (NREL)11. Those figures are the main source for the state of 2050. They in-
clude also small-scale PV and batteries. We assume and apply a general secu-
rity margin of 20%. To show the sensitivity, additionally lower costs assumptions
based on US’ studies with less conservative assumptions and for bigger installa-
tions are modelled.
Battery storage costs are currently still extremely high – especially for small stor-
age at individual houses. In future there is a big potential for cost reductions. Of-
ten market prices are at +20% compared to Germany in well working markets.
Storage / H2 prices have been updated based on EES 2021 conference (Oct. 21)
and on ATB figures as well as in an IEA report12. We assumed a mix of 30%
small installations (< 10 kW), 40% mid-sized installations (10–200 kW) and 30%
bigger installations (for Swiss conditions).
Costs of imported and exported electricity today is in the range of 5 cts/kWh (a bit
higher for exports as Switzerland gains some net income (SFOE, 2021). The
forecast for 2050 is almost impossible especially when taking into account the
turbulent situation on the electricity market during the last months. Generally,
higher costs are foreseen13. We assumed slightly higher costs for import (6
cts/kWh) and constant costs for export (5 cts/kWh), as Switzerland will tend to ex-
port more in summer and import more in winter in future based on the switch from
nuclear to PV.
11 https://atb.nrel.gov/electricity/2021/data
12 https://iea.blob.core.windows.net/assets/181b48b4-323f-454d-96fb-
0bb1889d96a9/CCUS_in_clean_energy_transitions.pdf
13 https://www.handelsblatt.com/politik/deutschland/co2-und-erdgaspreise-studie-strompreis-steigt-bis-2030-
um-50-prozent/27170486.html?ticket=ST-13976634-7m2L46hBf6kAVX9bDG0d-ap2
19
Table 2: Price assumptions for 2050. Round brackets: Lower costs based on
US studies. Square brackets: costs including share of agri-PV. For
PV and battery storage installation costs were used for modelling. For
non-optimized production types, the energy costs.
Nr Installation costs
in CHF/kW
Approx. energy costs
in cts/kWh
PV avg. on buildings 860 [786] (390) 6.9
Agri PV (farm land) 660 5.2
Battery storage10 330 (45) 9.2
Wind 11.0
Hydro 6.0 (mix of new and existing)
Hydrogen11 10.0
Gas power station (gas
and investment)
2000 CHF/kW 8.5
ETS 100 CHF/tCO2
Thermal electricity cost
incl. certification
11.1 – 16.8
Thermal electricity costs
based on H2 (e-fuels)
17.9 – 19.7
Imported electricity 6.0
Exported electricity 5.0
3.3.2 Six scenarios
Besides the main scenario based on ZERO Basis, five additional scenarios are
modelled in this project. They are defined as follows (Table 3):
Table 3: Six scenarios used in this study.
No. Scenario definition
1 100% renewable energy sources (RES) Switzerland with 0% net yearly import
and no additional import/export capacity restrictions of electricity to neighboring
countries
2 90% RES Switzerland with 10% net yearly import and no import/export capacity
restrictions of electricity to neighboring countries
3 90% RES Switzerland with 10% net yearly import and limited import/export ca-
pacities (3 GW) and with gas fired power plants (natural gas with carbon price of
60 CHF/t CO2)
4 90% RES Switzerland with 10% net yearly import and limited import/export ca-
pacities (3 GW) and with gas fired power plants (e-fuels)
5 84% RES Switzerland with 10% net yearly import/export, no import/export ca-
pacity restrictions and with 6% gas fired power plants (e-fuels)
20
No. Scenario definition
6 84% RES Switzerland with 10% net yearly import/export, no import/export ca-
pacity restrictions and with 6% gas fired power plants (e-fuels) with 13 GW of PV
on farm land
For all scenarios we also added the condition of isolating Switzerland and more
optimistic costs assumptions. Therefore, we modelled 24 different scenarios in to-
tal.
In Tables 4–6 today's system as well as six future scenarios are defined. To re-
duce complexity the renewable and non-renewable thermal production is com-
bined. Two additional major options were calculated:
1. Switzerland as an island with extremely limited transmission capacities to the
surrounding countries. Only in scenarios #3, #5 and #6 7 GW of import would
be allowed. Those scenarios are named with an “a” (e.g. “1a”).
2. Use of lower cost assumptions based on US studies. The source of the data
is the same (NREL ATB14), but the costs are for bigger systems and include
more optimistic outlooks.
Table 4: Annual electricity production in TWh 2018–2020 and 2050 with sce-
narios 1–6.
Type 2018–
2020
2050
Sc. 1
2050
Sc. 2
2050
Sc. 3
2050
Sc. 4
2050
Sc. 5
2050
Sc. 6
PV 2.17 33.6 30.0 30.0 30.0 27.0 27.0
Wind 0.14 4.3 2.15 2.15 2.15 2.15 2.15
Hydro 39.5 43.3 40.8 40.8 40.8 40.0 40.0
Nuclear 24.2 0 0 0 0 0 0
Net import 0.77 0 8.25 0 0 8.25 8.25
Therm. produc-
tion
3.0 3.1 3.1 11.35 11.35 6.9 6.9
Gross production 69.8 84.3 84.3 84.3 84.3 84.3 84.3
Net production 65.6 63.3 63.3 63.3 63.3 63.3 63.3
14 https://atb.nrel.gov/electricity/2021/data
21
Table 5: Installed capacities in GW 2018–2020 and 2050 with scenarios 1–6.
Opt. means: optimized in this project. Seasonal Hydro storage capac-
ity is in TWh.
Type 2018–
2020
2050
Sc. 1
2050
Sc. 2
2050
Sc. 3
2050
Sc. 4
2050
Sc. 5
2050
Sc. 6
PV 2.36 opt. opt. opt. opt. opt. opt.
Wind 0.14 2.2 1.1 1.1 1.1 1.1 1.1
Hydro (all types) 15.3 20.0 19.5 19.5 19.5 19.0 19.0
Nuclear 2.96 0 0 0 0 0 0
Therm. produc-
tion (all types)
0.97 0.97 0.97 3.75 3.75 2.25 2.25
Seasonal hydro
storage [TWh]
10 12 11 11 11 11 11
Table 6: Overview of scenarios 1–6.
Type 2018–
2020
2050
Sc. 1
2050
Sc. 2
2050
Sc. 3
2050
Sc. 4
2050
Sc. 5
2050
Sc. 6
Headline
E-Per-
spec-
tives
10% im-
port - no
gas
10%
gas
- re-
stricted
import
re-
stricted
import -
e-fuels
Import
and e-
fuels
Import,
e-fuels
and
Agri-PV
Share of re-
newables*
64% 99% 89% 89% 90% 84% 84%
Net annual
import
1% 0% 10% 0% 0% 10% 10%
Gas / E-Fuel
fired pp.
0% 0% 0% 11% 11% 6% 6%
* does not include share of imported H2 and share for air transport
3.3.3 Scenario #2 – 90% RES – 10% imported
The production of renewables is lowered (PV, wind, hydro) by 8.3 TWh to equal
10% of the electricity imported on an annual level. 10% are imported from neigh-
boring countries. Limitation: 7 GW for import, 10 GW for export.
Sensitivity test for: Variability of costs with 10% import (enhancing flexibility and
lowering price).
Background: The defined target for new renewables is relatively high. Therefore,
lowering the target indicates a more realistic option.
22
3.3.4 Scenario #3 – 90% RES – 10% based on gas fired power plants – limited im-
ports
The production of renewables is lowered (PV, wind, hydro) by 8.3 TWh. 10% are
produced with new gas fired power plants based on methane with ETS.
Import and export are still allowed, but limited to 3 GW.
Sensitivity test for: Variability of costs with gas fired power plants (duration of
usage) and with limitation to import and export of electricity.
3.3.5 Scenario #4 – 90% RES – 10% based on gas fired power plants run by re-
newable H2 – limited import
The production of renewables is lowered (PV, wind, hydro) by 8.3 TWh. 10% are
produced with new gas fired power plants based on renewable H2.
Import and export are still allowed, but limited to 3 GW.
Sensitivity test for: Variability of costs with 10% electricity produced with gas
fired power plants based on renewable liquids (H2 or methane). How much do the
costs vary when renewable liquids are used for the gas fired power plants?
3.3.6 Scenario #5 – 84% RES – 10% imported and 6% based on gas fired power
plants
The production of renewables is lowered (PV, wind, hydro) by 12.1 TWh to equal
approximately 84% of the electricity on an annual level. 10% are imported and
6% are produced with new gas fired power plants based on methane with ETS.
Sensitivity test for: Enhanced flexibility to 16% based on gas fired power plants
and import.
3.3.7 Scenario #6 – 84% RES – 10% imported and 6% based on gas fired power
plants – with option of agri-PV
The production of renewables is lowered (PV, wind, hydro) by 12.1 TWh com-
pared to scenario 1. 10% are imported and 6% are produced with e-fuels / re-
newable liquids (H2 or methane). PV on buildings is defined as 17 TWh and on
agricultural land 10 TWh.
Sensitivity test for: Enhanced flexibility with gas fired power plants and 10% im-
ports plus cost lowered by PV on agricultural land. What are the costs of not us-
ing PV on farm land?
Background: This scenario is assumed to lower the costs and will show a basic
cost level.
23
3.3.8 Overview of the scenarios
Figure 6 summarizes the contribution of all supply-side energy sources in each
scenario compared to the current situation. It clearly illustrates the central role to
be played by new firm PV generation, ranging from 35% of total generation in
scenarios #4 and #5 to 46% in scenario #1.
Figure 6: Supply-side electrical energy resources for all scenarios compared to
the current situation. The bottom part of the figure provides details for
the source labeled as ‘other’ in the top part. Note that scenario #4 is
the only scenario that does not include non-renewable (natural gas)
or possibly non-renewable (imports) resources.
0
20
40
60
80
100
Current Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6
Energy Generation TWh
Nuclear Hydro Other Firm Power PV Existing PV
24
3.4 Order of redispatch
The following Figure 7 shows the order of dispatch in the Clean Power Transfor-
mation (CPT) model for the basic scenario 1 not including any import / export.
PV capacity is deterministic and dependent on curtailment which is a driving in-
dependent variable. Optimization happens on curtailment / overbuilding in order
to minimized cost while respecting capacity and energy limits/ setpoints.
Figure 7: Dispatch model applied in the Clean Power Transformation (CPT)
model. PSH stands for pumped hydro storage.
We apply the CPT model to determine the optimum PV and battery resources
needed to meet demand firmly at the least possible cost while dispatchable re-
sources are optimally deployed toward this minimum cost/firm power generation
objective. The results of this optimization include the required quantities of new
battery storage, new PV, curtailed PV output (implicit storage), the electricity gen-
eration cost of the optimum supply-side/storage blend that will supply Swiss de-
mand 24x365.
Each meteorological year (2018, 2019 and 2020) is modelled alone to show the
sensitivity of inter-annual variations.
The annual (2020) dispatching of these resources is illustrated in Figure 8. 30-
day running means have been plotted to remove short-term fluctuations and im-
prove visualization. The top edge of the graph represents demand on the Swiss
25
grid. Note that the Swiss production is insufficient in winter and early spring, re-
quiring imports from the rest of Europe. However, production exceeds demand in
summer and is exported.
Figure 8: Annual dispatch of Swiss-based of supply-side resources for the year
2020. The top line of the stacked graph represents the Swiss grid
load
Net imports over the winter half year summed up to about 5 TWh during the last
20 years.
3.5 Climate change
We use a conservative approach as we do not include any climate change ef-
fects:
1. Climate change will enhance the run of hydro production in winter and lower it
in summer (a switch of about 0.6 TWh until 2050).
2. Climate change will lower the duration of winter. Therefore, the need for sea-
sonal storage of hydro is lowered.
3. Climate change will lower the heating needs – and enhances the cooling
loads (which will be much lower than the heating loads in 2050). Both would
be positive for integration of PV.
All three effects will lower the seasonal unbalance.
Nuclear
Biogas
Natural gas
PV
All hydro
Imports
Exports
MW
26
4 Results
4.1 General overview
The upcoming figures are mostly based on the meteorological year 2020. The
years 2018 and 2020 show very similar results, which is illustrated at the end of
this chapter. In Figure 9, we report the new PV capacity, curtailed PV output (im-
plicit storage), and battery storage required in each scenario to firmly meet de-
mand on the Swiss power grid.
Figure 9: New PV capacities (top), optimal curtailment (middle) and new battery
storage for scenarios 1–6 for connected (left) and stand-alone Swit-
zerland (right). Max. acceptable shows the assumed maximum of ac-
ceptable PV in Switzerland (55 GW ).
-
20
40
60
80
Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6 Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6
Net zero import/export Stand-alone Switzerland
NREL Technology Costs Swiss FOE Costs
NEW PV CAPACITY (GW)
0%
5%
10%
15%
20%
25%
30%
35%
40%
Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6 Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6
Net zero import/export Stand-alone Switzerland
NREL Technology Costs Swiss FOE Costs
OPTIMUM PV CURTAILMENT (%)
Max acceptable
-
20
40
60
80
100
Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6 Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6
Net zero import/export Stand-alone Switzerland
NREL Technology Costs Swiss FOE Costs
NEW BATTERY STORAGE (GWh)
27
New PV capacities (Figure 9 top) range from 33.5 GW (scenario 5 & 6 with net-
zero interconnectivity and optimistic technology costs) to 67 GW (scenario 1 with-
out interconnectivity and conservative costs). Applying optimistic cost assump-
tions reduces new PV requirements by about 9% overall compared to conserva-
tive costs. Operating the Swiss grid stand-alone would require 17% more PV to
be built than allowing net-zero interconnectivity. We plotted a “max acceptable”
line indicating the maximum amount of new PV that could be reasonably deployed
in the country. This amount is the result of a comprehensive analysis from the
Remund et al. (2019) that considered all deployable options (including roof space,
exclusion zones, farmland etc.) given current PV efficiencies. Importantly, all but
one scenario (#1a – autonomous grid) falls under this upper limit.
PV output curtailment (Figure 9, middle) ranges from 2% (scenario 5 & 6 with net-
zero interconnectivity and NREL costs) to 35% (scenario 1 without interconnectiv-
ity and Swiss FOE costs). Technology cost assumptions have a strong influence
on required curtailment. Applying optimistic cost reduces the need for it by an av-
erage of 41%. Stand-alone grid operation, without net-zero flexibility would in-
crease operational curtailment by 130%.
New battery storage requirements (Figure 9, bottom) range from 11.6 GWh (sce-
nario #6 with net-zero interconnectivity and conservative cost assumptions) to 85
GWh (scenario #5 and #6 with stand-alone grid and optimistic tech costs). Apply-
ing optimistic cost assumptions leads to two times more battery storage overall.
This significant difference is because future utility-scale NREL battery cost predic-
tions are very low compared to the conservative small-scale estimates (8 times
less) while the difference for PV between the two estimates amounts only to a fac-
tor of two. Interestingly, autonomous operation of the Swiss grid would only re-
quire 32% more battery storage than net-zero interconnected operation. In all
cases, required battery storage is low, amounting to 0.3 hours of full PV capacity
in the case of conservative cost assumptions, and ~1.2 hours in the case of opti-
mistic cost assumptions. The bottom line is that no new long-term storage is re-
quired beyond the small addition to the existing buffer hydro system (+10% / 1
TWh for scenarios 2-6, +20% / 2 TWh for scenario 1), as is often assumed when
envisaging ultra-high PV or wind penetration. This observation corroborates re-
sults obtained in the US (Perez, M., 2020). 10–85 GWh of batteries also seem
feasible compared to the expected electrical vehicle batteries, which will include
about 200 GWh of battery storage. Accessing 10% of this storage with by-direc-
tional loading systems would reduce the need of extra storage significantly.
28
Figure 10 reports the blended all-resources power generation LCOEs on the
Swiss power grid.
Figure 10: Swiss grid power generation costs for scenarios 1–6 and for con-
nected and stand-alone Switzerland.
Electricity production costs range from 5.2 cts/kWh (scenario 2 with optimistic
costs) to 8.6 cts/kWh (scenario 1 autonomous grid operation and conservative,
small-scale tech costs). Applying optimistic utility scale storage/PV cost assump-
tions reduces generation costs by an average of 22%. Importantly, as unlikely as
this configuration may be, stand-alone grid operation would only increase these
costs by an average of 7% i.e., not constituting a showstopper.
Figure 11 illustrates the critical role of implicit storage on the bottom line. Without
operationalizing PV overbuild and curtailment, production costs would be 71%
higher on average in the net-zero interconnected case, and 600% higher in the
stand-alone case.
The new annual dispatch of all resources is illustrated in Figure 11 for the 100%
RE (e-fuel) scenario #4. The top graph illustrates the net-zero import/export grid
configuration, while the bottom graph illustrates the autonomous grid configura-
tion. As in Figure 1, 30-day running mean have been plotted to remove short-
term fluctuations and improve visualization.
-
2
4
6
8
10
Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6 Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6
Net zero import/export Stand-alone Switzerland
NREL Technology Costs Swiss FOE Costs
SWISS GRID POWER GENERATION COST (¢/kWh)
29
Figure 11: Annual dispatch of supply-side resources for the year 2020 illustrated
for the 100% renewable scenario with e-fuels (#4). The top graph rep-
resents the net-zero interconnected configuration where winter im-
ports are energetically matched to summer export amounting to net-
zero. The bottom graph corresponds to the extreme stand-alone grid
configuration.
Biogas thermal
E-fuel thermal
Existing PV
All hydro
New PV direct to load
New PV via storage
Wind
MW
With import during winter
No import during winter
30
Figure 12 shows the share of energy production in scenario #6.
Figure 12: Share of energy production types for scenario #6 for 2050.
Electricity for battery charge and pumped hydro comes mostly from PV.
In Table 7 the main modelling results are concluded.
Table 7: Main results of the modelling for scenarios #1-#6 and #4a (no import).
Parameter Sc. 1 Sc. 2 Sc. 3 Sc. 4 Sc. 5 Sc. 6 Sc. 4a
PV installed capacity [GW] 50.1 41.0 41.0 41.0 36.6 37.0 48.1
PV curtailment [TWh] 7.9 4.7 4.7 4.7 4.1 4.5 11.1
LCOE [cts/kWh] 7.5 6.7 7.5 8.1 7.1 6.9 8.6
Battery Capacity [GWh] 24.8 19.8 19.9 19.9 11.9 11.6 26.6
Imports [TWh] 10.0 18.3 10.0 10.0 18.3 18.3 0.0
Overall, the results of the Energy Perspectives could be confirmed. The optimum
PV installation for this scenario (in this report #1) is 41 GW instead of the 37 GW
modelled in the perspectives including higher (14% instead of 9%) curtailment.
Results for all scenarios are given in Table 10 in the Annex.
31
4.2 Implicit storage impact
Figure 13 illustrates the importance of overbuilding and operationally curtailing
the PV resource on the bottom line: production costs would be an average of
63% higher across all scenarios for the net-zero interconnected configuration,
and 450% higher in the autonomous grid configuration. The main factor for this
cost difference is the amount of new battery storage required that would respec-
tively be 1300% and 7500% higher without PV oversize/curtailment.
Figure 13: Electricity production cost on the Swiss power grid as a function of PV
output curtailment for all scenarios. The top graph corresponds to the
interconnected grid configuration with net-zero import/exports with the
larger European grid. The bottom graph represents autonomous grid
configuration. Scenarios: #1: E-Perspectives 2050+; #2: 10% net im-
port; #3 and #4: 10% import, limited import capac., #5 and #6: 10%
import, 6% gas & agri-PV.
Net Zero
Import/Export
Autonomous
Grid
32
4.3 Sensitivity analysis: differences of meteorological years
The three years (2018–2020), analyzed independently, lead to very comparable
firm power production cost results overall as seen in in Figure 7 for the 100% re-
newable scenario #4.
Figure 14: Comparing 2018, 2019 and 2020 electricity production cost on the
Swiss power grid as a function of PV output curtailment for scenario
#4. The top graph corresponds to the interconnected grid configura-
tion with net-zero import/exports with the larger European grid. The
bottom graph represents autonomous grid configuration.
Net Zero Import/Export
Autonomous Grid
Curtailment (%)
LCOE (cts/KWh)LCOE (cts/KWh)
2018
2019
2020
33
4.4 Conclusions
Our investigation shows that high-RES solutions for Switzerland, with PV playing
a central role as a complementary resource to the country’s hydropower system,
are both physically and economically reasonable, despite the minor role wind
power can play, and the mediocre PV resource in winter months.
It is important to state that operational costs in all considered scenarios are rea-
sonable compared to current wholesale market prices in Switzerland (these have
been well above 20 cts/kWh the last couple of months15. The present ultra-high
RE costs are even reasonable when compared to earlier pre-crisis wholesale
prices (4–6 cts/kWh) noting that these earlier prices do not fully factor-in environ-
mental or strategic externalities which, as we see today with international ten-
sions, can be consequential.
Another particularly important observation is the result obtained for the 100%
RES scenario (#4). Not only are operational generation costs reasonable (6.5–
8.5 cts/kWh depending on technology and autonomy assumptions), but they
show the supply-side flexibility catalyst role that e-fuels can play, even as expen-
sive as they are expected to be at 18–20 cts/kWh.
Finally, we stress the importance of implicit storage (i.e., optimally overbuilding
the PV resources). Not implementing this deployment strategy would result in
higher prices on the network. It is therefore important to operationalize optimal
overbuilding and curtailment early-on, by e.g., implementing appropriate regula-
tions that would lead to firm power monetization, instead of current run-of-the-
whether PV production.
4.4.1 Costs of isolating Switzerland
The amount of imported net energy was defined for each scenario to lie between
18.3 TWh (scenarios #2, #5 and #6) and 0 (scenario #1, #3 and #4 with isolation
of Switzerland). Net imports are all taking place during the winter months. Table 8
shows the amount of electricity to be imported in 2050 for each scenario.
Table 8: Annual electricity imports in TWh with scenarios 1–6 with and without
isolation.
Net annual
import
Today Sc. 1 Sc. 2 Sc. 3 Sc. 4 Sc. 5 Sc. 6
With imports 5.0 10.0 18.3 10.0 10.0 18.3 18.3
Isolation (“a”) - 0.0 8.3 0.0 0.0 8.3 8.3
15 TNO & Fraunhofer ISE, (2022): Swiss Energy Charts. https://energy-
charts.info/charts/price_spot_market/chart.htm?l=en&c=CH&interval=year&year=2022&legen-
dItems=0000100000
34
Decoupling Switzerland from Europe increases costs by 5–15%. More PV, more
curtailment and more storage are needed. Only scenario #1a with no import
would induce PV capacities which are above the assumed threshold of 55 GW.
For a fully decoupled Switzerland with no additional imports of e-fuels (scenario
#1a) PV on farm land or significantly more hydro or wind energy would be
needed – all difficult to obtain due to political issues (landscape protection, food
production, biodiversity). Therefore, not mainly the costs, but the natural re-
sources and policy would be the main issues of decoupling Switzerland.
Just reducing imports (0% on annual bases and to 3 GW for power) and replac-
ing them by natural gas or e-fuels would induce only minor changes regarding
the costs. LCOE would rise slightly (about 15%). E-Fuels couldn’t be produced
largely in Switzerland but would have to be to imported (about 40% would be fea-
sible based on curtailment).
4.4.2 E-Fuels
As stated above, electricity based on e-fuels is modelled at a high cost of 18–20
cts/kWh. Nonetheless, scenarios #4–#6 including electricity from e-fuels also
show low overall LCOE costs (7–8 cts/kWh) because e-fuels have a low overall
share.
4–8 TWh of PV production would be curtailed optimally in these scenarios. This
could be used to produce e-fuels.
E-fuels are needed in the range of 9 TWh (scenario #4) and 5 TWh (scenarios
#5–#6). Regarding the round-trip efficiencies of 0.4, 40–50% of e-fuels could the-
oretically be produced in Switzerland. How many e-fuel will effectively be pro-
duced in Switzerland will depend on technology costs, transport costs, storage
available and costs of imported e-fuels.
4.4.3 Policy and Market
The lowest costs result with about 40 GW PV, 15% curtailment and 15 GWh bat-
teries, including a 10% rise of hydro power generation and storage (plus 1 TWh),
a rise in pumped hydro (from 2.9 to 5.7 GW) and an import of 5 TWh of e-fuels
(for electricity generation).
How to obtain this optimum is another question. The current policy and regulatory
framework most presumably will not induce enough investments to attain this:
With bigger shares, PV will start to cannibalize itself. At noon there will be more
PV than load and the prices will be zero or negative. Purely market-based mod-
els or power purchase agreements (PPA) will fail in this situation.
The electricity market in many Western countries and also in the EU and Switzer-
land is a copy of the market defined first in New England (USA) in the 1980s –
35
with no fluctuating renewable energies. It depends on marginal costs and the rule
of merit order. Many countries added an incentive for renewables and a capacity
market (Cramton et al., 2008) to reach more energy security as the energy only
market did not induce enough investments into additional capacities. This debate
is ongoing in Switzerland.
Germany, UK, Canada, USA and other major countries are targeting a 100%
RES based by 2035. As this production portfolio will include only to a small part
marginal costs a market based on marginal costs is at least debatable.
This shows the urgent need to develop new policies and new market models. A
short literature review (IRENA, 2017; Peng & Poudineh, 2017) indicates that
ideas exist, but the scientific foundation needs to be extended. Specifically, how
to secure overbuilding and thus minimize the overall costs is an open question.
Another question is how to regulate the curtailment: Who is doing this on which
level, based on which tools? IEA PVPS Task 1416 started a study to describe ex-
isting models. More work is needed.
4.4.4 Comparison to other studies
Several studies in Switzerland pointed out lately that the energy transition is not
easy to implement and that there are conflicting goals. The paper of Weiss et al.
(2021) about the “Energy Trilemma” showed that sustainability (CO2 emissions),
affordability (consumers’ costs) and security of supply are competing objectives.
Similar to this study, Thaler and Hofmann (2022) discussed the impossible en-
ergy trinity: energy security, sustainability and sovereignty.
In the paper about “Future Swiss Energy Economy” (Züttel et al., 2022) three ap-
proaches for the complete substitution of fossil fuels with renewable energy from
photovoltaics were considered: a purely electric system with battery storage, hy-
drogen, and synthetic hydrocarbons. This study noted that either huge areas for
PV or huge hydrogen storage or hydro power systems would be needed inducing
high costs and sustainability problems. Conflicting goals clearly exist: integration
in Europe, biodiversity, climate change and affordability of energy are competing
challenges to a certain level. However, Züttel et al. modelled unrealistic extreme
scenarios with 100% renewable energies (no imports also not for e-fuels) and no
efficiency gains – which in reality exists based alone on electrification for heating
and mobility and reduces the respective energy need by a factor of 2–3). In our
study based on Energy Perspectives 2050+, a part of the energy is imported
(28%) – PTL and e-fuels – and air transports aren’t included – to deliver those in
Switzerland would indeed be difficult.
Additionally, all three referenced papers did not include curtailment of PV. With
curtailment, a mostly isolated (with high security of supply) as well as e-fuels
based scenarios (with low CO2) lead to low costs of energy. As Table 10 shows,
16 https://iea-pvps.org/research-tasks/solar-pv-in-100-res-power-system/contacts_t14/
36
no optimum scenario for all objectives exists. Nevertheless, scenarios like #2a
(import of 8 TWh of electricity) and #4a (import of e-fuels, but not electricity)
would enhance electricity costs only marginally by 0.5 cts/kWh (to 8–8.5 cts/kWh)
– costs affordable for the Swiss customers.
The effects of higher levels of energy security (and less integration in the EU)
and climate protection is levelled out by higher PV installations and higher curtail-
ments. The energy trilemma exists, but is solvable to a big extent by overbuilding
PV which can be induced by suitable regulations and incentives.
4.5 Outlook
Four issues were not investigated in this study:
1. Nuclear power is not modelled as the study is based on the Energy Perspec-
tives 2050+. The newest nuclear power station in Switzerland Leibstadt was
built in 1984 and would be 66 years old in 2050, which is well above the
planned lifespan. Building new nuclear power plants is forbidden in Switzer-
land by the current law. Nevertheless, in our optimization model a new nu-
clear power station could be added. It could deliver the answer to how the
costs would change including a rather inflexible and expensive (15–20
cts/kWh) production method.
2. Alpine PV has not been included as well. PV installations in the Alps at alti-
tudes of 1500–2500 m above sea level with steep inclinations (e.g. 70° South)
would deliver almost the same electricity in winter as in summer. This would
ease integration. However, the potential is regarded as rather small (3–5
TWh) which would be quite small compared to the required 33 TWh – and
therefore would not change the seasonal distribution of the whole PV fleet sig-
nificantly.
3. The load was raised linearly by the factor of existing and foreseen energy
consumption (a growth from 70 to 84 TWh is modelled). No change regarding
the seasonal distribution has been made. Linearizing the load and not taking
into account that mainly winter electricity consumption will rise due to ex-
change of fossil heating systems with heat pumps, delivers too optimistic re-
sults regarding the winter load requirements.
4. The effects of climate change have been neglected. Neglecting climate
change induces conservative results. Climate change reduces seasonal ef-
fects: shorter and warmer winters, more precipitation in winter and less in
summer will be seen even if the Paris agreement limiting climate change to
1.5°C is reached.
5. Seasonal thermal storage isn’t modelled; this would ease the integration addi-
tionally.
37
The effects of the simplified load modelling and neglecting climate change could
level each other out to a certain degree.
38
5 References
Cramton P. & Stoft S. Forward reliability markets, 2008: Less risk, less market
power, more efficiency. Util. Policy 16, 194–201 (2008).
Figgener J. et al., 2020: The development of stationary battery storage systems
in Germany – A market review. J. Energy Storage 29, 101153.
IRENA, 2017: Adapting Market Design To High Shares of.
https://www.irena.org/-/media/Files/IRENA/Agency/Publica-
tion/2017/May/IRENA_Adapting_Market_Design_VRE_2017.pdf
O'Shaughnessy E., Cruce J. & Xu K., 2021. Solar PV Curtailment in Changing
Grid and Technological Contexts: Preprint. Golden, CO: National Renewa-
ble Energy Laboratory. NREL/CP-6A20-74176.
https://www.nrel.gov/docs/fy21osti/74176.pdf
Perez M., Perez R., Rabago K. & Putnam M., 2019: Achieving 100% Renewa-
bles: Supply-Shaping through Curtailment. PVTECH Power 2019, 19, 56–
61. Available online: http://www.pv-tech.org/ (accessed on June 11st
2021).
Perez M., 2020: Pathways to 100% Renewables across the MISO Region.
http://mnsolarpathways
Perez R., Perez M., Schlemmer J., Dise J., Hoff T. E., Swierc, Keelin A. P., Pierro
M. & Cornaro C., 2020: From Firm Solar Power Forecasts to Firm Solar
Power Generation an Effective Path to Ultra-High Renewable Penetration a
New York Case Study. Energies 2020, 13, 4489
Perez M., Perez R. & Hoff T., 2021: IMPLICIT STORAGE – Optimally Achieving
Lowest-Cost 100% Renewable Power generation. Solar World Congress.
International Energy Society.
Poudineh R. & Peng D., 2017: Electricity market design for a decarbonised fu-
ture. Oxford Institute for Energy Studies (OIES) Paper: EL 26.
https://www.oxfordenergy.org/wpcms/wp-content/uploads/2017/10/Elec-
trcity-market-design-for-a-decarbinised-future-An-integrated-approach-EL-
26.pdf
Remund J., Albrecht S. & Stickelberger D (2019). Das Schweizer PV-Potenzial
basierend auf jedem Gebäude. in Photovoltaik Symposium Bad Staffel-
stein.
SFOE, 2021: Swiss Federal Office of Energy, 2021: Swiss Energy Perspectives
2050+. https://www.bfe.admin.ch/bfe/en/home/policy/energy-perspectives-
2050-plus.html
39
Thaler P. & Hofmann B., 2022: The impossible energy trinity: Energy security,
sustainability, and sovereignty in cross-border electricity systems. Polit. Ge-
ogr. 94, 102579.
Tong D., Farnham D.J., Duan L. et al., 2021: Geophysical constraints on the reli-
ability of solar and wind power worldwide. Nat Commun 12, 6146.
https://doi.org/10.1038/s41467-021-26355-z
Züttel A. et al., 2022: Future Swiss Energy Economy, 2022: The Challenge of
Storing Renewable Energy. Front. Energy Res. 9, 1–17.
Weiss O., Pareschi G., Georges G. & Boulouchos K., 2021: The Swiss energy
transition: Policies to address the Energy Trilemma. Energy Policy 148,
111926.
40
6 Annex
Table 9: Input definitions.
Energy production & import in TWh
Type 2018-2020 2050 Sc. 1 2050 Sc . 2 2050 Sc . 3 2050 Sc. 4 2050 Sc. 5 2050 S c. 6
PV 2.17 33.6 30 30 30 27 27
Wind 0.14 4.3 2. 15 2.15 2.15 2. 15 2.15
Hydro 39.5 43. 3 40.8 40.8 40. 8 40 40
Nuclear 24.2 000000
Import 0.77 08.25 0 0 8.25 8.25
Therm. production 33.1 3.1 11.35 11.35 6.9 6.9
Gross produc tion 69.8 84.3 84.3 84. 3 83.4 84.3 84. 3
Net producti on 65.6 63.3 63. 3 63.3 63.3 63. 3 63.3
Check sums
Total
69.78 84.3 84. 3 84.3 84. 3 84.3 84.3
New renewables
539.8 34.05 34. 05 34.05 31. 05 31.05
All renewables
44.5 83.1 74. 85 74.85 74.85 71. 05 71.05
Reduced renewables
8.25 8.25 8. 25 12.05 12.05
Gas fired pp
9.45 9.45 5 5
Import (annual share)
1% 0% 10% 0% 0% 10% 10%
Share of CH EE
64% 99% 89% 89% 90% 84% 84%
Installed GW
Type 2018-2020 2050 Sc. 1 2050 Sc . 2 2050 Sc . 3 2050 Sc. 4 2050 Sc. 5 2050 S c. 6
PV 2.36
50 41 41 41 37 37
Wind 0.14 2.2 1.1 1. 1 1.1 1.1 1. 1
Hydro 15.3 20 19. 5 19.5 19.5 19 19
Nuclear 2.96 000000
Import
Therm. production 0.97 0.97 0.97 3. 75 3.75 2.25 2. 25
Scenario de finition 2050 Sc. 1 2050 Sc . 2 2050 Sc. 3 2050 Sc. 4 2050 S c. 5 2050 Sc . 6
Headline
E-
Perspec tives
10% import -
no restric tions
No import -
gas
No import - e-
fuels
Import and e-
fuels
Import, e-fuels
and agri-PV
Share of renewables 64% 99% 89% 89% 90% 84% 84%
Net annual import 1% 0% 10% 0% 0% 10% 10%
Import restri ctions no (10 GW) no (10 GW) no (10 GW) yes (3 GW ) yes (3 GW ) no (10 GW) no (10 GW )
Share of gas fired pp. 0% 0% 0% 11% 11% 6% 6%
Thermal prod. [GW ] 0. 97 0.97 0.97 3. 75 3.75 2.25 2.25
Thermal prod. [TWh] 3.1 3.1 3.1 11.35 11.35 6.9 6. 9
Thermal prod. Renew. Share 61% 61% 61% 17% 17% 28% 28%
Fuel costs 2050 Sc. 1 2050 S c. 2 2050 Sc . 3 2050 Sc . 4 2050 Sc. 5 2050 Sc. 6
Headline
E-
Perspec tives
10% import -
no restric tions
No import -
gas
No import - e-
fuels
Import and e-
fuels
Import, e-fuels
and agri-PV
Natural gas 30 30 30 30
Emis sion rate [ t CO2/MW h] 0.4 0.4 0. 4 0.4
Efficiency (gas -> electricity) 0.6 0. 6 0.6 0.6 0.6 0. 6 0.6
Power stat ion (invest., o&m)
[CHF/ MWh]
35 35 35 35 35 35 35
CO2 emiss ion certificates
[CHF/ tCO2]
60 100 100 100
CO2 removal / sequest ration
[CHF/ tCO2]
150 150 150
E-Fuel (green H2)
[CHF/ MWh]
100 100 100
Total natural gas / certif
[CHF/ MWh]
100 111 111 141
Total natural gas / sequestr
[CHF/ MWh]
124 124 168
Total E-Fuel (H2)
[CHF/ MWh]
197 179 179
Natural gas wit hout certif 85 85
Rene w. Costs 2050 Sc. 1 2050 Sc . 2 2050 Sc. 3 2050 Sc. 4 2050 S c. 5 2050 Sc . 6
Headline
E-
Perspec tives
10% import -
no restric tions
No import -
gas
No import - e-
fuels
Import and e-
fuels
Import, e-fuels
and agri-PV
PV ins tall. Costs
[CHF/ MWh]
860 860 860 860 860 786
PV prod. Cos ts [cts /kWh] 6. 9 6.9 6.9 6.9 6. 9 6.3
Wind [ cts/ kWh] 11 11 11 11 11 11
Hydro [cts/kWh] 6.04 5.80 5. 80 5.80 5.80 5. 80
Battery install cos ts
[CHF/ MWh]
330 330 330 330 330 330
Battery [cts/kWh] 9.2 9.2 9. 2 9.2 9.2 9.2
Import [cts/kWh] 5.0 6. 0 6.0 6.0 6.0 6. 0 6.0
Export [cts/kWh] 5.0 5. 0 5.0 5.0 5.0 5. 0 5.0
41
Table 10: Average results of modelling based on 2018, 2019 and 2020 meteor-
ological years. Imports will happen during winter time (imports and
exports are listed also in Table 9).
Costs Scenari o PV installed PV curtailed Battery capac. LCOE Thermal prod. Imports
[GW] [%] [ GWh] [ cts/kWh] [TWh] [TWh]
Swiss 150.1 14% 24.8 7.46 3.1 10
Swiss 241.0 11% 19.8 6.73 3.1 18.3
Swiss 341.0 11% 19.9 7.53 11.35 10
Swiss 441.0 11% 19.9 8.13 11.35 10
Swiss 536.6 11% 11.9 7.09 6.9 18.3
Swiss 637.0 12% 11.6 6.91 6.9 18.3
Swiss 1a 67.8 35% 34. 8 8.60 3. 1 0
Swiss 2a 48.0 24% 26. 4 7.22 3. 1 8.3
Swiss 3a 48.1 24% 26. 6 8.02 11. 35 0
Swiss 4a 48.1 24% 26. 6 8.62 11. 35 0
Swiss 5a 39.9 17% 19. 5 7.38 6. 9 8.3
Swiss 6a 40.2 18% 20. 3 7.19 6. 9 8.3
USA 145.9 9% 45.9 5. 53 3.1 10
USA 237.4 3% 36.0 5. 16 3.1 18. 3
USA 337.4 3% 36.2 5. 96 11.35 10
USA 437.4 3% 36.2 6. 55 11.35 10
USA 533.5 0% 85.0 5. 76 6.9 18. 3
USA 633.5 0% 85.0 5. 75 6.9 18. 3
USA 1a 62.6 30% 55. 6 5.99 3.1 0
USA 2a 43.9 18% 44. 7 5.34 3.1 8.3
USA 3a 44.0 18% 45. 4 6.13 11. 35 0
USA 4a 44.0 18% 45. 4 6.73 11. 35 0
USA 5a 36.6 11% 35. 2 5.85 6.9 8.3
USA 6a 36.6 11% 35. 2 5.85 6.9 8.3