Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Integration of Renewables in the Swiss Energy System


Published on

Integration of Renewables in the Swiss Energy System

Published in: Data & Analytics
  • Be the first to comment

  • Be the first to like this

Integration of Renewables in the Swiss Energy System

  1. 1. WIR SCHAFFEN WISSEN – HEUTE FÜR MORGEN Integration of Renewables in Swiss Energy System Evangelos Panos :: Energy Economics Group:: Paul Scherrer Institute Kannan Ramachandran :: Energy Economics Group:: Paul Scherrer Institute 72th Semi-annual ETSAP meeting, 11-12th December, Zurich
  2. 2. Overview of the Swiss Energy System, 2015 Page 2 0 20 40 60 80 Solar Wind Other Nuclear Hydro ELECTRICITY GENERATION (TWh) 0 5 10 15 20 25 Solar Wind Other Nuclear Hydro ELECTRICITY CAPACITY (GW) 0 200 400 600 800 1000 Heat RES Wood Electricity Other Gas Oil Coal FINAL ENERGY CONSUMPTION (PJ) 0 10 20 30 40 Power generation Services and agriculture Industry Residential Transport ENERGY RELATED CO2 EMISSIONS (Mt) Swiss energy strategy 2050: aims at gradually phasing out nuclear and promoting renewables and demand side efficiency 56% 38% 13.7 3.3 Solar
  3. 3. • We study integration measures for variable renewable generation (VRES) from wind and solar PV in Switzerland for the horizon 2015 – 2050, such as:  Reinforcing and expanding the grid network  Deploying local storage, complementary to pump hydro, like batteries and ACAES  Deploying dispatchable loads such as P2G, water heaters and heat pumps • The study was performed in the context of the ISCHESS project, which is a collaboration between the Paul Scherrer Institute and the Swiss Federal Institute of Technology (ETH Zurich), funded by the Swiss Competence Center Energy and Mobility (CCEM) Objectives of the research Page 3
  4. 4. • High intra-annual resolution with 288 typical hours (3 typical days, 4 seasons, 24h/day) • In this research, the model includes:  Higher detail in the electricity sector, variability in the RES generation, ancillary services and power plant dispatching constraints  But, it excludes oil transport and has aggregate representation of industry Methodology – The Swiss TIMES Energy Systems Model (STEM) Page 4 Swiss TIMES Energy system Model (STEM) Fuel supply module Fuel distribution module Demand modulesElectricity supply module Resource module Electricity import Uranium Natural gas Hydrogen Electricity export Electricity Gasoline Diesel Renewable · Solar · Wind · Biomass · Waste Electricity storage Hydro resource · Run-of rivers · Reservoirs CO2 Demand technologies Residential - Boiler - Heat pump - Air conditioner - Appliances Services Industires Hydro plants Nuclear plants Natural gas GTCC Solar PV Wind Geothermal Other Taxes & Subsidies Fuel cell Energy service demands Person transportat ion Lighting Motors Space heating Hot water Oil Transport Car fleet ICE Hybrid vehicles PHEV BEV Fuel cell Bus Rail Macroeconomicdrivers(e.g.,population,GDP,floorarea,vkm) Internationalenergyprices(oil,naturalgas,electricity,...) Technologycharacterization(Efficiency,lifetime,costs,…) Resourcepotential(wind,solar,biomass,….) Biofuels Biogas vkm-Vehicle kilometre tkm-tonne kilometre LGV-Light goods vehicles HGV-Heavy good vehicles SMR-steam methane reformer GTCC-gas turbine combined cycle plant Oil refinery Process heat FreightsTrucks HGV Rail Natural gas Heating oil Other electric
  5. 5. • Different grid levels, characterised by transmission costs and losses • A set of power plants can be connected to each level:  Power plants are differentiated by costs, efficiency, technical constraints & availability • A linearised approximation of the Unit Commitment problem is formulated Representation of the electricity sector in STEM Page 5 Very High Voltage Grid Level 1 Nuclear Hydro Dams Imports Exports Distributed Power Generation Run-of-river hydro Gas Turbines CC Gas Turbines OC Geothermal High Voltage Grid Level 3 Large Scale Power Generation Medium Voltage Grid Level 5 Wind Farms Solar Parks Oil ICE Waste Incineration Large scale CHP district heating Wastes, Biomass Oil Gas Biogas H2 Low Voltage Grid Level 7 Large Industries & Commercial CHP oil CHP biomass CHP gas CHP wastes CHP H2 Solar PV Wind turbines Commercial/ Residential Generation CHP gas CHP biomass CHP H2 Solar PV Wind turbines Lead-acid batteries NaS batteries VRF batteries PEM electrolysis Pump hydro CAES Lead-acid batteries NaS batteries VRF batteries Li-Ion batteries NiMH batteries PEM electrolysis Lead-acid batteries Li-Ion batteries NiMH batteries
  6. 6. + 4 nodes for nuclear power plants • With a reduction algorithm from FEN/ETHZ we map the detailed transmission grid to an aggregated grid with 𝑁 = 15 nodes and 𝐸 = 319 lines, based on a fixed disaggregation of the reduced network injections to the detailed network injections Representation of electricity transmission grid Page 6 MAPPING −𝐛 𝐝 ≤ 𝐇 𝐝 × 𝐠 𝐝 − 𝐥 𝐝 ≤ 𝐛 𝐝 𝐇 𝐝 is the PTDF matrix of the detailed network, 𝐃 is the fixed dissagregation matrix,𝐠 𝐝 , 𝐠 𝐚 are vectors with elelectricity injections, 𝐥 𝐝 , 𝐥 𝐚 are vectors of electricity withdrawals, and 𝐛 𝐝 , 𝐛 𝐚 are vector of line capacities; superscripts d=detailed, a=aggregated network The matrix 𝐃 is not unique, since there are infinite ways in which an aggregate injection can be distributed between multiple nodes; here, it allocates power injections according to the original distribution of generation capacity in the detailed power flow model −𝐛 𝐚 ≤ 𝐇 𝐝 × 𝐃 × 𝐠 𝐚 − 𝐥 𝐚 ≤ 𝐛 𝐚 DETAILED NETWORK AGGREGATED NETWORK
  7. 7. • To capture variability, we need the variance of the mean availability of wind and solar per typical hour in STEM • By bootstrapping a 20-year meteorological sample, we calculate the distribution of mean availability per typical hour in STEM • We move ± 3 in this distribution to introduce variability of RES per typical hour Representation of stochastic RES variability Page 7 Bootstrapped Distribution of Mean Photovoltaic Capacity Factors (availability): Summer (left), Winter (right) • Storage capacity must accommodate downward variation of the Residual Load Curve (RLDC) and upward variation of non-dispatchable generation • The dispatchable peak generation capacity (incl. storage) must accommodate upward variation of the RLDC and downward variation of non-dispatchable generation Source: Singh, A. PSI, 2016
  8. 8. • Power plants commit capacity to the reserve market based on their operational constraints and shadow price of electricity generation and reserve provision • In each of the typical hours the demand for reserve is endogenously calculated from the joint probability distribution function (p.d.f.) of the individual p.d.f. of forecast errors of supply and demand  The sizing is based on both probabilistic and deterministic assessment  We move ± 3 s.d. on the joint p.d.f to estimate the reserve requirements Ancillary services markets – provision of reserve Page 8 The standard deviation of the joint probability distribution of forecast errors is : 𝜎2 𝑠𝑜𝑙𝑎𝑟 + 𝜎2 𝑤𝑖𝑛𝑑 + 𝜎2 𝑙𝑜𝑎𝑑 And the reserve requirement in hour 𝑡 is given as: By assuming uncorrelated random variables: 𝑅 𝑡 = 3 ∗ 𝜎2 𝑠𝑜𝑙𝑎𝑟 ∙ 𝐺 𝑡 2 𝑠𝑜𝑙𝑎𝑟 + 𝜎2 𝑤𝑖𝑛𝑑 ∙ 𝐺 𝑡 2 𝑤𝑖𝑛𝑑 + 𝜎2 𝑙𝑜𝑎𝑑 ∙ 𝐿 𝑡 2 + 𝑎 ∙ 𝑃 𝑚𝑎𝑥 𝐺𝑖 generation from option 𝑖 , 𝐿 load, 𝑃 𝑚𝑎𝑥 largest grid element
  9. 9. A range of “what-if” scenarios was assessed along three main dimensions: 1. Future energy policy and energy service demands, assuming nuclear phase out by 2034:  Energy service demand growth: low growth vs high growth  Electricity imports: allow net imports vs autarky  Climate change mitigation policy: mild policy vs 70% reduction in CO2 emissions in 2050 from 2010 levels 2. Location of new gas power plants and installed capacity as % of the total national capacity 3. Grid expansion: allowing grid reinforcement beyond the “Grid 2025” plans of Swissgrid or not in total about 100 scenarios were assessed with the STEM model based on the Cartesian product of the above combinations across the three dimensions Long term scenarios analysed with STEM Page 9 Corneux (NE) Chavalon (VS) Utzenstorf (BE) Perlen (LU) Schweizerhalle (BL) Case 1 20.0 20.0 20.0 20.0 20.0 … … … … … … Case 11 0.0 33.3 33.3 33.3 0.0 … … … … … … Case 26 33.3 33.3 0.0 0.0 33.3
  10. 10. • Electricity consumption increases 4 – 30% from 2015 ( 0.1 – 0.8% p.a) • New gas power plants replace existing nuclear capacity • Under climate policy VRES could provide up to 28% of the domestic supply Electricity consumption continues to increase and gas, VRES & imports replace nuclear by 2050 Page 10 -10 0 10 20 30 40 50 60 70 80 2015 2050 (max) 2050 (min) ELECTRICITY GENERATION & CONSUMPTION IN 2050 (TWh) The results correspond to ranges among the 100 scenarios assessed
  11. 11. • The requirements for secondary reserve almost double in 2050 from today’s level and peak demand shifts from winter to summer • Hydrostorage is the main contributor, followed by gas turbines (after 2040) • Flexible CHPs and batteries could participate in the reserve provision by forming “virtual plants”, especially in those scenarios with high reserve requirements Electricity consumption continues to increase and gas, VRES & imports replace nuclear by 2050 Page 11 SECONDARY RESERVE MAXIMUM DEMAND AND PROVISION IN 2050 (MW) 0 100 200 300 400 500 600 700 800 2015 2050 (max) 2050 (min) Maximum contribution per technology The results correspond to ranges among the 100 scenarios assessed
  12. 12. • High shares of VRES require electricity storage peak capacity of ca. 30 – 50% of the installed capacity of wind and solar PV (together) • Above 14 TWh of VRES generation there is a significant storage deployment • About 13% of the excess summer VRES production is seasonally stored in P2G (~ 1 TWh) Storage needs increase with VRES deployment Page 12 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 8 10 12 14 16 18 20 22 24 Installedstoragecapacity(GW) TWh of wind and solar PV electricity production Pump hydro (pumping capacity) Batteries (4h max discharge) P2G (only as seasonal storage) • Small scale batteries are driven by distributed solar PV installations • Medium scale batteries are driven by wind and large scale PV and CHP • Large scale batteries complement pump hydro storage when it is unavailable Max. Total battery capacity 5.5 GW Large scale batteries 0.5 GW Small scale batteries 3 GW Medium scale batteries 2 GW WIND AND SOLAR PV ELETRICITY VS STORAGE CAPACITY Each data point in the graph corresponds to a different long term scenario and year
  13. 13. Contribution of each sector to the total electricity stored in water heaters and heat pumps Residential 85-97% Industry <2% Services 3-23% • Electricity storage in water heaters and heat pumps could account for 8 – 24% of the total electricity consumption for heating • Accelerated deployment of electric load shifting when electricity consumption for heat exceeds 13 TWh • Large potential for load shifting is in water heating (resistance heating) followed by space heating in buildings Dispatchable loads help in easing electricity load peaks in the stationary end-use sectors Page 13 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 8 13 18 Electricitystoredinwater heatersandheatpumps (TWh) Electricity consumption for heating (TWh) ELECTRICITY STORED IN WATER HEATERS AND HEAT PUMPS VS ELECTRICITY CONSUMPTION IN HEATING IN 2050 (TWh) Each data point in the graph corresponds to a different long term scenario
  14. 14. • Restrictions in grid expansion results, due to congestion, in:  Higher system costs, up to 90 BCHF (+10% when grid expansion is allowed) in 2020-50  Non-cost optimal deployment of electricity supply options  Less electrification of demand and reliance on fossil-based heating The system-wide benefits from the electricity grid expansion outweigh the costs Page 14 1100 1200 1300 1400 1500 1600 1700 No grid expansion Grid expansion No grid expansion Grid expansion Reference Stingent climate policy Max Min CUMULATIVE UNDISCOUNTED ELECTRICITY AND HEAT SYSTEM COST, BHCF/yr, 2020 – 2050 The results correspond to ranges among the 100 scenarios assessed
  15. 15. • Grid expansion results in cost savings, mainly in heating sectors, directly (e.g. technology change via heat pumps) and indirectly (e.g. less costs for imported fuels)  Under a stringent climate change mitigation policy and high demand, these costs could be up to 3 BCHF/yr  About 1/3 of the savings occur in the electricity supply and 2/3 in the stationary sectors The system-wide benefits from the electricity grid expansion outweigh the costs Page 15 -3500 -3000 -2500 -2000 -1500 -1000 -500 0 500 1000 Electricity power plants Electricity T&D Net imports of electricity and fuels Heat supply Total Max Min ELECTRICITY AND HEAT SYSTEM COSTS SAVINGS DUE TO GRID EXPANSION, MCHF/yr, 2020-2050 The results correspond to ranges among the 100 scenarios assessed
  16. 16. • Electricity consumption continues to increase by 0.1 – 0.8% p.a., and especially after 2030 due to new electricity uses; it could reach over 70 TWh/yr by 2050 • VRES can contribute up to 24 TWhe (or 28% of the domestic supply) after the nuclear phase-out • Storage peak capacity investments about 30 – 50% of the installed wind and solar PV capacity; beyond 14 TWhe accelerated deployment of storage is inevitable • About 13% of the excess electricity production from VRES in summer is seasonally stored in P2G pathways, driven by the high seasonal electricity price differences under stringent climate change mitigation scenarios • Water heaters and heat pumps could contribute in easing electricity peaks and they could shift 8 – 24% of the electricity consumption for heat Conclusions Page 16
  17. 17. • Grid reinforcement results in net economic benefits for the whole electricity and heat supply system of Switzerland on the order of 0.5 – 3.0 BCHF/yr. over the period 2030 - 2050 • Rebound effects prevail in grid reinforcement, and a social planner shall prioritize the expansion of those lines with high benefit/cost ratio for the energy system (insights are provided form the dual values of the grid constraints in the model) • Further work is needed to overcome some important limitations, such as:  Regional representation also for the heat supply and not only for electricity  Consideration of N-2 grid security constraints  More detail in technical representation of storage technologies (e.g. depth of discharge, ramping rates, max discharging times, etc. ) Conclusions ctd. Page 17
  18. 18. Page 18 Wir schaffen Wissen – heute für morgen Thank you for your attention. Evangelos Panos Energy Economics Group Laboratory for Energy Systems Analysis Paul Scherrer Institute
  19. 19. • Without batteries and grid, there is 30 – 50% less deployment of wind and solar electricity compared to the case when both options are available  Batteries are important for the integration of VRES to cope with their variability  Grid expansion is important to integrate large amounts of VRES production ( >16 TWh) • Total system costs can be 10 – 14% higher if both batteries and grid expansion are unavailable  In particular climate policy costs could increase by more than 50% (from 103 to 160 BCHF) Storage and grid expansion are required to realise the VRES potential and lower climate policy costs Page 19 0 5 10 15 20 25 Reference Climate policy TWhofwindandsolarPV No Batteries, No grid expansion With Batteries, Grid expansion IMPACT OF BATTERIES AND GRID EXPANSION IN THE DEPLOYMENT OF WIND AND SOLAR PV POWER 1100 1150 1200 1250 1300 1350 1400 1450 Reference Climate policy Undiscountedcumulative costBCHF No Batteries, No grid expansion With Batteries, Grid expansion IMPACT OF BATTERIES AND GRID EXPANSION IN THE TOTAL SYSTEM COSTS Results on the left graph corresponds to ranges; results on the right graph is for a scenario with high demand and electricit y imports