Ea Energianalyse

Integration af vindkraft

Viking Link og andre tiltag for integration af vind

1 Denne rapport er udarbejdet af Ea Energianalyse a/s for .dk, november 2015.

Indhold 1. Introduktion og sammenfatning – Metode – Resultater 2. Overblik over elmarkedet 3. Elementer i integrationsomkostningen 4. Modellering og forudsætninger 5. Centralt scenario – Kapacitetsudvikling – Nøgleparametre • Prisudvikling • Vindens værdi 6. Integrationstiltag

Udarbejdet af: Hans Henrik Lindboe, Jesper Werling, Nina Ea Energianalyse a/s har udarbejdet rapporten som uafhængig konsulentvirksomhed og anser indholdet i denne rapport for at være Dupont og János Hethey velfunderet. Aktører bør dog basere sig på deres egne vurderinger, når resultater fra denne rapport anvendes. Fremskrivninger er i deres natur behæftet med stor usikkerhed og baseret på interne og eksterne antagelser på fremtidige udviklinger, som vil falde Ea Energianalyse anderledes ud i realiteten. Frederiksholms Kanal 4, 3. th. 1220 København K Ea Energianalyse a/s kan ikke gøres ansvarlig for tab som følge af anvendelse af indholdet i denne rapport. T: 88 70 70 83 F: 33 32 16 61 E-mail: [email protected] Web: www.eaea.dk

2 Siden midten af 1990’erne, er vindkraften i Danmark steget fra at udgøre knap 5% til ca. 33% af den indenlandske elforsyning i 2013, og ca. 40% i 2015. Energiaftalens målsætninger om at nå 50% inden 2020 vil ifølge Energinet.dk’s forudsætninger for udbygning af vindkraft blive overopfyldt, og udviklingen ventes at fortsætte frem mod 2035, hvor vindkraft vil udgøre op til 84% af det klassiske elforbrug. Klassisk elforbrug er eksklusiv elforbrug til fjernvarme, individuelle varmepumper og elbiler. Samtidig viser de nationale planer for vedvarende energi en markant VE udbygning i landene omkring os. Elproduktionen fra vindkraftanlæg afhænger af, hvor meget det blæser, og det er velkendt, at udfordringerne ved at integrere Introduktion vindkraften i elsystemet stiger efterhånden som vindkraftandelen øges. Integration af vindkraft stiller krav om et fleksibelt elsystem med tilstrækkelige kapacitet til, at Denne rapport er udarbejdet af Ea Energianalyse for Energinet.dk efterspørgslen efter elektricitet også kan dækkes med høj sikkerhed når det ikke blæser. Et udtryk for omkostningen ved at integrere vindkraft er vindmøllernes lavere afregningspris i markedet end gennemsnittet for de øvrige elproducenter. Energinet.dk har bedt Ea Energianalyse om at analysere effekten af forskellige tiltag til at integrere vindkraft, herunder Vikingforbindelsen, varmepumper og andre tiltag. Endvidere ses der på tiltagenes interne afhængighed.

En konklusion er, at Vikingforbindelsen og store varmepumper hver for sig har positiv samfundsøkonomi, samt at de to vindintegrationstiltag ikke væsentligt påvirker hinanden økonomisk. Værdien af at gennemføre begge tiltag er således større end blot at gennemføre ét af tiltagene. Det betyder at en ”både og” løsning kan være en økonomisk attraktiv løsning i forhold til at integrere vindkraft på lang sigt.

Kilde: Energistyrelsen

3 Beregninger og analyser er hovedsageligt udført med el- og varmemarkedsmodellen Balmorel. Modellen omfatter en repræsentation af el- og fjernvarmesystemet i Norden og store dele af Europa. Modellen beregner bl.a. produktion, transmission og elpriser baseret på forudsætninger for udviklingen af brændselspriser, udbygning med vedvarende energi og andre vigtige parametre.

Analysen er struktureret omkring et basis scenario, der beskriver en sandsynlig udvikling for el- og Metode fjernvarmesystemerne. I dette centrale scenarie beregnes udbygningen af produktionsapparatet i en simulering med aggregeret tidsopløsning. For at analysere integrationen af vindkraft gennemføres efterfølgende beregninger på timeniveau, og samtlige resultater for drift og økonomi af systemet i denne Modelberegninger rapport er baseret på timeberegningerne. Som væsentligste indikator for integration af vindkraft sammenlignes markedsværdien af vindkraft-el med markedsværdien af gennemsnits-el. Indikatoren for om et tiltag er attraktivt er, at tiltaget har positiv samfundsøkonomi for den samlede region.

Scenario Scenarioanalyse (I parentes angives forkortet scenarienavn) (Ændringer i forhold til Basis scenario) • Sandsynlig udvikling af el- og fjernvarmesystemerne under hensyntagen til planlagt og aftalt energipolitik. Tilgangen svarer i Basis scenario store træk til tankegangen bag IEAs New Policy Scenario, som (Base) anvendes i den årlige publikation World Energy Outlook (WEO). Basisscenariet beregnes uden Viking kablet • Etablering af Viking Link i 2022. Produktionsapparatet fastholdes og (VikingLink) er derfor det samme som i basisscenariet. • Gradvis indfasning af store varmepumper i det danske Udbygning med store varmepumper fjernvarmesystem. Forsyner 20 % af fjernvarmebehovet i 2030 (Heatpumps) (modeloptimeret)

Øget fleksibilitet på større kraftværker • Installation af elpatroner/dampturbinebypass i 2020 på de større (CPP_flex) kraftværker i Danmark. I alt 3.500 MW varme (ikke modeloptimeret) Øget fleksibilitet på individuelle varmepumper • Øget fleksibilitet på elforbrug i individuelle varmepumper i form af (Indiv_heatpumps) mindre varmelager til ca. 4 timer. Kombination af Viking Link og udbygning med • Udvikling af produktionsapparat som i varmepumpescenariet. store varmepumper (Combi 1) • Etablering af Viking Link i 2022 Kombination af Viking Link og øget fleksibilitet • Udvikling af produktionsapparat som i CPP_flex på større kraftværker (Combi 2) • Etablering af Viking Link i 2022

Kombination af Viking Link, udbygning med • Udvikling af produktionsapparat som i varmepumpescenariet + 4 store varmepumper og øget fleksibilitet på CPP_flex større kraftværker (Combi 3) • Etablering af Viking Link i 2022 Basisscenariet viser med de valgte forudsætninger en kraftig omstilling af det overordnede energisystem, hvor VE får en stigende betydning og udgør 2/3 af den samlede produktion i 2040 mod ca 1/3 i 2014. Fluktuerende VE fra vind og sol står for ca. 40 % i 2040, mens andelen i Danmark når op over 65 %.

På kort sigt betyder de stagnerende brændsels- og CO2-priser, at elpriserne i 2020 er på niveau med 2014- priserne (2015-priserne er lavere, bl.a. pga. vådår). Efter 2020 forventes stigende priser på baggrund af Resultater stigende CO2- og brændselspriser. Efter 2030 fører de yderligt stigende mængder VE til, at elpriserne ikke stiger yderligere på trods af en fortsat stigning i el- og brændselspriser. Der forventes dog et større antal Basis scenario timer med hhv. lave og høje elpriser. Beregningsår 2014, 2020, 2030 og 2040. Som følge af de øgede mængder vindkraft, stiger afregningsprisen for vindkraft ikke i samme grad som de gennemsnitlige elpriser. I Vestdanmark forventes den gennemsnitlige afregningspris for on- og offshore vindkraft at ligge mere end 25 % under den gennemsnitlige spotpris på langt sigt (prispres). Inden 2030 viser modelresultaterne et prispres på under 20%. Det bemærkes dog, at modelværktøjet sandsynligvis undervurderer prispresset, da modellen har fuld information om produktions- og forbrugsforhold (ingen afvigelser mellem planer og faktisk drift). For 2014 viser modellen fx en afvigelse på omkring 8% mod realiserede ca. 12 %.

Elpriser og vindafregningspriser (Vestdanmark) 600

] 500

400

[DKK/MWh 300

Elpris 200

100

0

Statistik El-pris Statistik vindafregningspris Estimeret El-pris Estimeret vindafregningspris 5 Scenariernes effekt på integrationen af vindkraft vises ved Scenarierne har indflydelse på ændringer i prispresset på vind-el. Prispresset er defineret som produktionsmønstrene i det samlede el- og forskellen imellem vindafregningsprisen og den gennemsnitlige fjernvarmesystem. De væsentligste effekter er: spotpris i Danmark målt i procent. Der fokuseres her på scenariernes effekt for vindintegration, men det understreges, at Viking Link tiltagene også tjener andre formål. Således anvendes Viking Link • ikke kun til eksport og import af vindkraft, og de store Kort sigt: Reduceret produktion fra naturgas i Resultater varmepumper leverer fjernvarme, hvorved der fortrænges Storbritannien. Øget produktion fra kulkraft i anden varmeproduktion. Viking Link reducerer prispresset med andre lande. Integration af vindkraft knap 4 %-point ved introduktionen i 2022 samtidig med at den • Længere sigt: Mindre nedregulering af gennemsnitlige spotpris øges . På længere sigt reduceres vindkraft (mindre curtailment) prispresset med 2-2.5%-point. De store varmepumper får først Varmepumper for alvor effekt på længere sigt (2040), med reduktion af • Reduceret produktion fra biomasse KV i prispresset på mere end 4%-point. Danmark. Øget produktion fra kulkraft og naturgas i andre lande. I kombinationsscenarierne svarer reduktionen ca. til summen af • Besparelser på omkostninger til de respektive enkelte scenarier (Viking Link, varmeproduktion i DK. varmepumpescenariet og Fleksibilitet på større • Længere sigt: Mindre nedregulering af kraftvarmeværker) . Elpatroner/turbinebypass på de større vindkraft kraftværker i Danmark har størst effekt på kort sigt, hvor Fleksibilitet på større kraftvarmeværker prispresset på vind reduceres med 7%-point. Små individuelle • varmepumper har stort set ingen effekt på prispresset. Reduceret produktion fra biomasse KV i Danmark. Øget produktion fra brunkul og 30% naturgas i andre lande. (Mindre udtalt end i varmepumpe scenariet)

• Mindre nedregulering af vindkraft * * [%] 20% • Besparelser på omkostninger til

vind varmeproduktion i DK.

på Fleksibilitet på individuelle varmepumpe 10% • Begrænset effekt på det overordnede system.

Prispres • Mindre nedregulering af vindkraft 0% 2010 2020 2030 2040 Base VikingLink Heatpumps CPP_flex Indiv_heatpumps Combi 1 Combi 2 Combi 3

*For 2014 vises statistisk data

6 Scenarierne viser generelt en positiv samfundsøkonomi for 4000 det samlede system – set over en længere årrække. De samfundsøkonomiske fordele er ikke alene baseret på 3000 forbedret integration af vindkraft, men inkluderer også ændrede produktionsomkostninger i el- og 2000 fjernvarmesystemet generelt. Ved at sammenligne de Resultater forskellige scenarier kan konkluderes, at en kombination af 1000 Viking Link og varmepumper er en økonomisk fordelagtig Samfundsøkonomi strategi for at integrere vindkraft på langt sigt. Udbredelsen 0 af varmepumper i Danmark og opførelsen af VikingLink er 2020 2022 2030 2040 tiltag der begge virker positivt på vindintegration uden at

Årlige besparelse Årlige besparelse [mio. DKK/år] -1000 det ene tiltag forringer økonomien i det andet væsentligt. VikingLink Store varmepumper Fleksibilitet store kraftværker Fleksibilitet indiv. VP For især Viking Link, udbygning med store varmepumper og Kombinationsscenarie 1 Kombinationsscenarie 2 kombinationen af disse to scenarier, er den Kombinationsscenarie 3 samfundsøkonomiske gevinst størst på længere sigt (2030 og frem), mens der på kort sigt kan være et Scenario Nutidsværdi samfundsøkonomisk tab forbundet med scenarierne. Den [mia. DKK] beregnede samfundsøkonomi gælder dog for den samlede Viking Link 5.6 region, og det er forskelligt hvordan enkelte lande påvirkes. Dette gælder især for Viking Link scenariet, da Udbygning med store varmepumper 20.2 transmissionsudbygning påvirker landende forskelligt. For Øget fleksibilitet på større kraftværker 5.1 Danmark er samfundsøkonomien af Viking Link positiv Øget fleksibilitet på individuelle igennem hele perioden, mens der på kort sigt kan opstå et 1.3 samfundsøkonomisk tab for andre lande. Dette skyldes bl.a. varmepumper Kombination af Viking Link og landenes forskellige produktionsmix samt den særlige 22.8 udbygning med store varmepumper engelske pris på CO -emissioner. 2 Kombination af Viking Link og øget 7.6 fleksibilitet på større kraftværker Samfundsøkonomien for scenariet med øget fleksibilitet på Kombination af Viking Link, udbygning individuelle varmepumper afhænger af omkostningen for at med store varmepumper og øget 25.3 realisere scenariet, som især vil omfatte udgifter til fleksibilitet på større kraftværker individuelle varmelagre og styring. Disse omkostninger er *Der er ikke inkluderet eventuelle omkostninger til individuelle ikke inkluderet her. varmelagre m.m.

7 BACKGROUND ELECTRICITY PRICES

8 The Nordic Power Market

Nord Pool Spot Market The Nord Pool Spot daily competitive auction at 12:00 establishes a price for each hour of the next day (24 hours). Prices are created for each bidding area, as well as an overall ‘system price’. A bid states prices for each hour and corresponding volumes in a specific area. Area prices differ from the ‘system price’ due to congestion of limited transmission capacity. The DK1 area represents Western . The system price (SYS) is the market clearing price which disregards transmission capacity. SYS is the basis for financial power derivatives. Nord Pool covers the Nordic Countries (-Iceland) and the Baltic States.

700 600 500 Continental connections 400 300 The Nordic power market is interconnected with continental Europe. DKK/MWh 200 Market based flows between Nord Pool and Northwest European markets result from iterative price calculations - market coupling. 100 Absent of congestion, bids from a generator in Norway essentially compete - with bids from a power plant in France. 1 6 11 4 9 2 7 12 5 10 3 8 1 6 11 4 9 2 7 12 5 10 3 8 1 6 11 4 9 2 7 12 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 20142015 DK-West Germany Avg. system price (nordpool) Data source: Nordpoolspot.com. Nominal prices.

9 Historical Western Danish Spot Price

700 High CO2 and fuel prices Dry 600 Financial crisis: Declining demand

Collapse of CO2 and 500 Dry year fuel prices Rather wet and Dry year Rather warm (low 400 dry demand in Wet Nordics)

300 DKK/MWh 200

100

- 1 4 7 10 1 4 7 10 1 4 7 10 1 4 7 10 1 4 7 10 1 4 7 10 1 4 7 10 1 4 7 10 1 4 7 10 1 4 7 10 1 4 7 10 1 4 7 10 1 4 7 10 1 4 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 DK-West Germany Avg. system price (nordpool) Data source: Nordpoolspot.com. Nominal prices. 100 Hydrological conditions have a significant impact on the Nordic price formation. 80

• In dry years, Sweden, Finland and Norway increase net-imports to compensate for (%) lack of hydro generation. 60 • In wet years abundance of hydro allows plant owners to lower the prices of their 40 supply offers. 20 Besides the availability of hydro power, the main driver of short-term movements in the Reservoir filling Reservoir filling 0 power price are fuel prices and the price of CO2.

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 20132014 Norwegian hydro reservoir storage levels in percent of capacity.

10 Average Prices by generator type

Captured prices vary by generation technologies Simple time-weighted averages provide a first indication of the market potential for various generation technologies to capture prices, but since wholesale prices vary hour-by-hour, average quarterly and yearly prices differ between technologies. • Dispatchable generation with high short-run marginal costs capture higher (albeit fewer) prices on average. • Technologies with low short-run marginal costs capture lower prices, but have more operating hours. • Intermittent generation generally captures lower prices, particularly when the resource (e.g. wind) is simultaneously abundant across a wide region. Solar power generally has an advantage of coincidence between generation and high demand. This will erode with a significant increase in penetration. Danish power generators are primarily large power stations, decentralised CHPs and wind 700 turbines. Recent years have also see an 600 increase in distributed solar PV. 500 Since 2002, prices captured by generators in 400

Western Denmark have in relation to the DKK/MWh 300 average hourly price been: 200 • Central power stations +8% 100 • Decentralised CHP +9% 0 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 • Wind power -10% 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 • Solar power +5% DK-West: Time average DK-West: Central DK-West: Decentral DK-West: Wind DK-West Consumption Calculations based on data from Energinet.dk. Nominal prices.

11 BALMOREL MODEL

12 The Balmorel Model

Balmorel is used for analysing electricity, CHP and heat in internationally integrated markets. Applications are Modelling power and district heating systems long-term planning, as well as detailed short-term or operational analysis. • Hourly dispatch optimisation The model can be used to calculate generation, • Optimal investments in generation and transmission transmission and consumption of power and heating • Price formation in partial equilibrium price formation on hourly basis as well as to optimize the electricity, allowing for market and stakeholder analysis. heat and transmission capacity in the system. Prices are generated from system marginal costs, emulating optimal competitive biding and clearing of the market.

Areas of application include: Main model inputs • Existing generation capacities, respective unit’s technical and • International power market development economic data, investment options (incl. refurbishments) and • Analyses of wind integration technology development. • Security of electricity supply • Transmission system infrastructure, and options and costs for • The role of demand response capacity expansion. • The role of natural gas • Projected demand for power and heating • Expansion of electricity transmission • Projected fuel and CO -prices. 2 • Markets for green certificates • Policies, taxes and support schemes • Electric vehicles in the power system • Environmental policy evaluation

13 Representation of the international power market

Transmission grid 2014 Modelling power transmission and price areas

Development in the international and interconnected power and energy system has significant implications on the development in any singular price area. • The analysis described in this report includes calculations in the regions shown on the map. Most relevant are Denmark and Great Britain (note that Northern Ireland is included in the Ireland price area and not in Great Britain), the Nordic countries, Germany, France the Netherlands and Belgium. • Individual countries are subdivided into regions, between which the most significant power transmission congestions occur. • In the Nordpool countries, these regions coincide with the price zones in Nordpool. Presently, the German power market has only one price zone (together with Austria), in spite of congestion in the internal grid. Modelling Germany as one price region without consideration of internal congestion in Germany would lead to unrealistic power flows and export opportunities, e.g. for Danish power plants however, therefore 4 price regions are modelled for Germany.

14 Representation of Denmark's district heating and CHP

Individual and aggregated district heating systems In Denmark, there are approximately 420 different district heating networks, with numerous plants and heat producing units contained in these systems. 11 centralised areas account for approx. 62% of the total district heating demand, roughly 10% is located in medium scale areas (more than 1 PJ annual heat demand), and 28% are located in small decentralised heating areas. The centralised and medium sized areas are represented individually, while the small areas are aggregated according to the type of supply. However, all areas partially supplied by municipal solid waste are represented individually, regardless of their size.

Units connected to centralised heating areas in DK

Apart from district heating units and systems, commercial CHP 120 plants supplying local industry heat demands are also included in 100 two separate areas in Western and Eastern Denmark. 80 Data for existing power plants and district heating demands are based on annual statistics reported by all producers to the Danish 60 Energy Agency for 2011. 40 Forecasts for heat demand factor in both energy savings and (PJ/year) 20 expansion of the respective district heating networks are based on analyses of the Danish district heating systems by Ea Energy 0

Analyses and COWI for the Danish Energy Agency during 2013. 2013 2020 2025 2035 Net district heating heating demand district Net Small areas Medium scale Centralised areas Units connected to medium scale DH systems individually represented

15 ASSUMPTIONS AND MODEL SETUP

16 Power consumption development

Nordics: Power demand is based on national forecasts where available. Power Demand Nordics and North-West Europe Otherwise, forecasts are based on National Renewable Energy Action Plans (NREAPs) until 2020.Development after 2020 is based on a BASREC-study with input from participating countries. 700 Denmark: Development for Denmark based on assumptions from the Danish TSO Energinet.dk (Includes moderate introduction of EVs and individual heat pumps) 600 Great Britain: Based on DECC projection (Updated energy and emissions projections 2014) excluding Northern Island and Auto-generators. According to the DECC projection electricity usage increases by roughly 23% between 2020 500 and 2035, mainly due to higher demand in the residential and service and agriculture sector, in parts due to declining impact of current efficiency policies. Other EU: Statistic data based on ENTSO-E, forecasts based on statistic data and 400 the trend in EU Energy, Transport and GHG Emissions – Trends to 2050 by the European Commission Power demand Denmark 300

50.000 Power Power demand[TWh] 40.000 200

30.000

20.000 100 [GWh] 10.000 0 0 2010 2015 2020 2025 2030 2035 2040 Annual Annual consumption* power 2015 2017 2019 2021 2023 2025 2027 2029 2031 2033 2035 Classical consumption Individual heatpumps BE DE FR NL Electric vehicles Electrification of railway NO SE DK GB *source: Energinet forudsætninger, May 2015; including losses; excluding demand for large heat pumps and electric boilers (determined endogenously by Balmorel)

17 Power transmission capacities

Development of the overall transmission grid is based on the Ten Year Transmission expansion between Network Development Plan 2014, developed by the transmission 2014 and 2040 system operators within ENTSO-E. The planned capacities are supplemented information by TSOs, where updated information is available. Significant changes for Denmark (excluding the Viking Link). • Nordlink 1.4 GW (DE-NO) by 2020 • CobraCable 0.7 GW (DK-NL) by 2020 • Kriegers Flak adds transmission option between Eastern DK and DE by 2019 (not shown on map) • German internal grid, based on the TSOs’ latest grid development plan (NEP2014), scenario B. Data until 2020 was directly implemented; expansion beyond 2020 appeared optimistic with regards to the controverse ongoing discussions. Therefore, the most controversial expansion corridors were assumed to be delayed to 2025. Transmission capacity between North West DE and Southern DE increases significantly between 2020 and 2025. Futher expansion beyond 2025 between South and North DE • Danish internal grid 0.6 GW (DK East-DK West) by 2030

18 Fossil fuel price development

140 Oil prices reclining from bottom The oil market was bottomed out at the beginning of 2015 having 120 reached the lowest level since the onset of the financial crisis. Tight oil plays in the US, OPEC’s reluctance (inability) to curb supplies - 100 accompanied with weak demand - created a significant shift in the supply-demand balance. US inventories have only recently started to decline, putting upward pressure on the forward curve, indicating a 80 modest recovery.

60 Natural gas In recent years, shale gas developments, especially in the US, have led to 40 a glut in the LNG markets and a trend towards a decoupling of natural

gas and oil prices has depressed natural gas prices on European trading Fuel price price Fuel (DKK2015/unit) 20 hubs. Replication of the US shale gas has so far not been very successful. In spite of pressure from new technology bringing additional reserves on- - stream, and stagnant growth, eventually global competition for access to

energy commodities and pricing by scarcity is likely to recur.

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 Year Note: FWD prices are extended to be constant in real terms beyond observations. Coal prices at historic lows Historical - Coal IEA - Coal FWD - Coal Convergence - Coal Coal prices are also at historic lows. Cheaper natural gas in the US, Historical - Crude oil IEA - Crude oil increase of renewable energy and weak power demand in the EU and FWD - Crude oil Convergence - Crude oil Historical - Natural gas IEA - Natural gas not least a stagnation of Chinese coal consumption growth are key FWD - Natural gas Convergence - Natural gas reasons for the slide of coal prices. Coal prices are now lower than at the IEA price points bottom of the financial crisis. Prices are projected to converge from today’s towards the IEA’s main projection between 2020 and 2030.

19 CO2-emissions price development (EUA & Denmark) Driving the green transition 1,20 350

) At current levels, the CO2-emission prices alone cannot drive the green transition of the 2 1,0 300 European Energy Systems. CO2-futures reflect the current price of EUA’s (EU emission allowances). These can be banked from year to year (free storage). ,80 250 200 ton CO ,60 The IEA WEO 2014 states CO2-prices in the EU-ETS of 20 USD/ton in 2020 rising to 40

150 USD/ton in 2040 (in real 2013 values). By comparison, the EU impact assessment assumes DKK/

,40 [ 100 a 40 EUR/ton CO2-price by 2030 to attain the 40% CO2 reduction target. Adding energy ,20 50 saving measures reduces the 2030 price to 22 EUR/ton and adding support for RE price deployment and EE reduces the price to 11 EUR/ton. In a conservative scenario without - - 2

enabling polices for technological development the 2030 CO2-price is 53 EUR/ton if the CO same reduction target has to be met. The key question is if the CO2-price will be brought on track, or if the RE-transition will be driven by subsidies. Year Historical - CO2 Price IEA - CO2 Price The base assumption on prices for EUA’s in this report follows forward prices to 2020 and FWD - CO2 Price Convergence - CO2 Price after this it is assumed that policy will drive prices to converge towards the IEA WEO 2014 New Policies scenario, with prices intersecting in 2030. This corresponds to assuming a Note: FWD prices are extended to be constant in real terms beyond observations.

continued mix between supporting RE by subsidies and a driving CO2-price, with an ] 2 1.000 emerging importance of the CO2-price.

CO In 2013, the UK introduced a carbon price floor, aiming at ensuring a minimum CO -price 800 2 of £16/tCO2 in 2013 to £70/tCO2 in 2030 (2009-prices). The system is implemented by 600 adding a national tax on CO2 (CPS, Carbon Price Support), which should cover the

DKK/ton difference between the EU ETS price and the total target. However, in 2014 the UK [ 400 introduced a cap on the national tax of £18/tCO2 for 2016-2019 to limit the difference between the EU ETS price and the total national CO -tax. Currently, no policies for such a

price 2

2 200 cap beyond 2019 are in place, and in principal CPS will rise again after 2019. However, in

CO this report it has been assumed, that the national tax will be phased out between 2020 0 and 2030, and thus the effective carbon price in the UK will be the same as in the rest of Europe as of 2030. The reasoning behind this is twofold: 1) It is not desirable to base the economic value of an interconnector between Denmark and the UK on a difference in the Years CO2-regulation system, which by its nature is uncertain. 2) With increasing connection of EU ETS (DKK2015/ton) UK CPS (DKK2015/ton) the power systems in the UK and continental Europe, it becomes less likely, that a large UK total (DKK2015/ton) UK total target (DKK2015/ton) difference in the effective CO2-price will be acceptable for power producers in international competition.

20 Scenario setup INTEGRATION MEASURES

21 Scenario setup

• Baseline investment: Model opimized power and heat capacity in all modelled countries. • Base: Hourly baseline run, dispatch optimization. • Viking Link: Identical to the Base scenario, but with an added 1.4 GW transmission link between Great Britain and Denmark. • Heat pumps: Generation capacities of the Baseline investment for all countries but Denmark. Capacities in Denmark are optimized by the model after lowering the tariff on electricity for district heat generation to a value of 150 DKK/MWh. • Flexible CPP: Generation capacities from base scenario, but installation of electric boilers at larger CPP plants in Denmark. Some of this effect could be achieved by using turbine bypass. • Individual heat pumps: Generation capacities from base scenario, but increased flexibility of individual heat pumps. Assumed option to move demand by approximately 4 hours. • Combi 1: Identical to the Heat pumps scenario but with added Viking Link. • Combi 2: Identical to Flexible CPP but with added Viking Link • Combi 3: Identical to Heatpumps scenario but with added Flexible CPP and Viking Link

Baseline Investment run

Combi 1 Combi 2 Combi 3 Individual Base Viking Link Heat pumps Flexible CPP Heat pumps + CPP flex + Heat pumps + Heat pumps Viking Link Viking Link CPP flex + Viking Link

22 Model results BASELINE INVESTMENT RUN

23 Investment approach

In the Baseline investment run the electricity and heat capacities are optimized by the model. The table below shows the general approach with respect to which generation is included exogenously and which generation capacity is decided on by model optimization.

Renewable capacity Other capacity

Denmark Exogenous until 2035. Investment beyond Decommissioning + Investments

Nordics Exogenous until 2030. Investment beyond Decommissioning + Investments. Nuclear Exogenous Grat Britain Exogenous until 2035. Investment beyond Decommissioning + Investments. Nuclear Exogenous Germany Exogenous until 2035. Investment beyond Decommissioning + Investments. Nuclear Exogenous Other EU Exogenous until 2020 (NREAP). Investment beyond Decommissioning + Investments. Nuclear Exogenous Transmission Exogenous until 2030 (TYNDP). Constant beyond

24 Capacity development

Great Britain Denmark 180.000 Solar In Denmark, investments are made endogenously for 160.000 Wind thermal capacity based on the current tax and subsidy 140.000 Biogas schemes. Capacities for wind and solar power are given Waste exogenously up to 2035 and based on Energinet.dk 120.000 forecasts. However, due to the recent price development of Biomass 100.000 solar power, the generation of one larger offshore wind farm Pumped hydro has been replaced by solar power. 80.000 Hydro 16.000 Power Power capacity[MW] 60.000 Other fossils 14.000 40.000 Natural gas Solar 20.000 Coal 12.000 Wind 0 Nuclear 10.000 Biogas 2014 2020 2022 2030 2040 8.000 Waste Renewable capacity is modelled exogenously until 2035. Biomass 6.000

Development of nuclear capacity is based on the average of Other fossils Power Power capacity[MW] the Reference scenario and the Existing policies scenario of 4.000 Natural gas the Updated Energy and Emission projections 2014 (UEP 2.000 Coal 2014) by the Department of Energy and Climate Change (DECC). Investment and decommissioning decisions for fossil 0 fuels are modelled endogenously. As a result all coal capacity 2014 2020 2022 2030 2040 is decommissioned by 2020.

25 Capacity development

Nordics and North-West Europe 3000 • Nuclear generation for France and Belgium includes plans for phase-out (by 2025 for Belgium) and decrease in capacity (50% of 2500 generation by 2025 in France)

Solar • Development in Germany is determined by the

2000 Wind so-called Energiewende, which aims for 80% of ]

Biogas the generation to be supplied by renewable [GW Waste energy in 2050. The assumptions in the present 1500 Biomass analysis are based on the analysis framework Hydro from the German network development plan for

Power Power capacity Other fossils 1000 2015. Nuclear power is phased out by 2023. Lignite • Norway: Build out of wind and expansion of Natural gas hydro. 500 Coal • Sweden: Closure of two nuclear blocks at Ringhals Nuclear and two blocks at Oskarshamn until 2020. Constant nuclear capacity beyond 2020. 0

Increasing capacity of wind.

2020 2022 2030 2040 2020 2022 2030 2040 2020 2022 2030 2040 2020 2022 2030 2040 2020 2022 2030 2040 2020 2022 2030 2040 BELGIUM FRANCE GERMANY HOLLAND NORWAY SWEDEN

26 Generation Great Britain renewable Overall generation generation 3500 Solar Assumptions for solar generation in the UK are based on the National grid’s Future Energy Scenario (FES) - Slow progression. The biomass and biogas 3000 Wind capacities are based on the Updated Energy and Emissions projections report Biogas 2013 (UEP 2013), generation in the actual model runs can be lower than 2500 mentioned in the source if applied subsidies aren’t high enough to ensure Waste sufficient amount of full load hours. Onshore/Offshore wind capacities are based on the UEP 2013 generation and scaled to fit the total generation of the 2000 Biomass UEP 2014 . Maximum onshore wind (2040) the model is allowed to invest in is Hydro based on FES gone green scenario. All RE capacities have been reduced with a 1500 2,4% factor to account for capacity in Northern Ireland, which is not included in Other fossils

the model runs. Power generation[TWh] Comparison of sources for development of RE in the UK 1000 Natural gas Lignite Biogas 300 500 Coal 250 Biomass 200 Nuclear Waste 150 0 100 Renewable 2014 2020 2022 2030 2040 50 Other Solar 00 The overall generation in the modelled part of Europe shows a trend of RE generation generation [TWh] RE Offshore wind decreasing fossil fuel generation, compensated by a growth in renewable

Onshore wind generation - especially wind and solar power.

BaseInvest BaseInvest BaseInvest BaseInvest BaseInvest

UEP 2014

FES Slow progression Slow FES progression Slow FES progression Slow FES progression Slow FES progression Slow FES 2014 2020 2022 2030 2040

27 Wind generation in Europe

1.200 Wind power generation in Europe doubles in the modelled area between 2014 and 2020. By 2040, wind 1.000 generation as increased to 547% fo the 2014 value. A large share of that is wind

power generation in Great Britain, ] 800 France and Germany. The large share of TWh wind power creates challenges for the system to ensure a high value of the generated wind power. Curtailment of 600

generation [ 547% wind power occurs, but is less than 2%

by 2040. wind

400 Total 390%

200 249% 220%

100%

0 2014 2020 2022 2030 2040 OTHER FR DE GB BALTICS NORDICS DK

28 Model results BASE SCENARIO

29 Electricity prices and prices captured by wind power Denmark 500 The base scenario shows stable electricity prices until

2020, mainly attributed to the stable fuel and CO2-prices. 450 After 2020, increasing fuel- and CO2-prices lead to 400 increasing average electricity prices in Denmark until 2030. After 2040, electricity prices do not increase

350 further, even though fuel and CO2-prices continue to increase. This is explained by the increased generation 300 from renewable energies, leading to more hours with 250 lower prices, thereby compensating for the higher electricity prices at times, when fossil fuel fired power 200 plants set the market price. Thus, even though average

Electricityprice [DKK/MWh] electricity prices stay stable after 2030, the price profile 150 shows more volatile prices. 100

50 Prices captured by wind power are approximately 18% below the average electricity price by 2020. This gap 0 increases to almost 29% by 2040, when the share of wind in Denmark and in the overall system is highest. The Statistic El-price DK_W Statistic wind price DK_W increasing volatility after 2030 is also reflected by an Statistic El-price DK_E Statistic wind price DK_E increasing gap from 2030 to 2040. El-price DK_W Wind price DK_W El-price DK_E Wind price DK_E

30 Electricity prices

Comparison DK_W and GB The decrease in price difference can be partly explained by the decreasing difference in CO2 prices. The UK has a carbon price 500 160 support (CPS) of currently £18/ton CO2 added to the EU-ETS 450 140 carbon prices. As of 2030, the CPS is assumed to be zero, resulting 400 in the same CO prices for the UK and Europe. A continuation of 120 2 350 the current level of the CPS would result in increasing electricity 300 100 prices in Great Britain, thereby increasing the price difference to Denmark and increasing the value of a link between the two 250 80 regions for both Denmark and Great Britain from a merchant line 200 60 perspective.

150 [DKK/MWh] 40 Depending on the marginal emission factor in continental Europe 100 and Great Britain, the CO2-price difference of around 200 DKK/ton Electricityprice [DKK/MWh] 20

50 in 2020 can explain differences in the electricity price of around Absolute difference and DK GB 0 0 85 DKK/MWh. All else equal, the price difference would therefore 2014 2020 2022 2030 2040 be reduced by this value by 2030. Base - DK_W Base - GB Base - Diff 350 300 By 2020 the difference between average electricity prices in Denmark and Great Britain is large (about 140 DKK/MWh). As 250 the electricity price in Denmark rises towards 2030 the 200 difference with British prices decreases over the years. 150 Between 2030 and 2040 the Danish price remains constant while 100 a drop in British prices lowers difference between average 50 electricity prices further. Carbonprice [DKK/ton] 0

DK GB

31 Duration curves for electricity prices

950  Great Britain

750 The number of negative prices depends – amongst others – on the subsidy schemes for variable 550 renewable generation. In the model runs, a simplified assumption on a constant RE subsidy has 350 been applied. Subsidy schemes are expected to evolve over the years to not pay a subsidy at times 150 with zero or negative prices (as it is the case for offshore already). In this -50

1 case, fewer hours with negative than shown on the

Electricityprice [DKK/MWh]

293 585 877

2921 8177 1461 1753 2045 2337 2629 3213 3505 3797 4089 4381 4673 4965 5257 5549 5841 6133 6425 6717 7009 7301 7593 7885 8469 1169 duration curves will occur. However, prices will still -250 2020 2022 2030 2040 be close to zero, as the negative prices reflect situations with excess generation. West Denmark  950 The duration curves for the electricity prices in both 750 Denmark and Great Britain show a trend towards 550 more hours with both higher and lower electricity prices. A difference between average prices can be 350 one important driver for connecting regions. However, since the low and high prices occur at 150 different times in Denmark and Great Britain, -50

connecting the regions can be beneficial even 1

Electricityprice [DKK/MWh]

585 877

though the price difference between average 293

1461 1753 2045 2337 2629 2921 3213 3505 3797 4089 4381 4673 4965 5257 5549 5841 6133 6425 6717 7009 7301 7593 7885 8177 8469 -250 1169 electricity prices descreases between 2014 and 2040. 2020 2022 2030 2040

32 Price difference DK West and GB duration curve Base scenario

1.500

1.000

500

0

1

231 461 691 921

1611 4141 1151 1381 1841 2071 2301 2531 2761 2991 3221 3451 3681 3911 4371 4601 4831 5061 5291 5521 5751 5981 6211 6441 6671 6901 7131 7361 7591 7821 8051 8281 8511

-500 Electricity price [DKK/MWh] price Electricity

-1.000 2020 2022 2030 2040

Number of Number of Average of In the base scenario without the Viking Link, the price is initially hours higher hours higher absolute price higher in Great Britain for virtually all the hours of the year. In later GB price DK_W price difference years this distribution becomes more balanced. The average of the 2020 8469 197 148.3 absolute price difference decreases between 2020 and 2030, then 2022 8204 387 110.8 rises drastically in 2040, where Great Britain shows lower prices for more than 3000 hours. 2030 6081 2260 98.4 2040 5109 3356 168.8

33 Model results VIKING LINK SCENARIO

34 Viking Link’s generation impact in Europe

Overall changes Germany 3,5 6 Solar 3 Wind Wind 4 Biogas Biogas 2,5 Hydro 2 2 Waste Waste Biomass 1,5 Biomass 0 Other fossils Hydro [TWh] 1 [TWh] Lignite Other fossils -2 0,5 Natural gas Lignite 0 Coal -4 Natural gas Nuclear

Changein power generationVikingLink -0,5 Changein power generationVikingLink Coal -6 -1 Nuclear 2022 2030 2040 2022 2030 2040

The introduction of the Viking Link in the system has an impact on the generation patterns across Europe. The biggest absolute changes are seen in Great Britain with a reduction in natural gas and biomass, especially in 2022. Initially, the Viking link results in higher coal generation in Germany. This effect is reversed by 2040, where several countries, among which Great Britain, Denmark and Germany show less wind power curtailment due to better integration. Changes in hydro generation are due to the model setup, and are accounted for in the economic calculations.

35 Viking Link’s generation impact

Great Britain Denmark Introduction of the Viking Link leads to changes in the overall 0,90 electricity generation. In Great Britain, a large decrease of natural 0,80 gas generation is seen especially for the year 2022. Biomass generation decreases to a lesser extent in 2022 and 2030. By 0,70 2040, the improved options for electricity export lead to a Wind 0,60 reduction of wind curtailment and increased generation from nuclear power. 0,50 Biogas 0,40 2 [TWh] Biomass 0,30 1 Wind Natural gas 0 Biogas 0,20 Coal -1 Biomass 0,10 Hydro Changein power generationVikingLink -2 - Other fossils -0,10 -3 Natural gas

[TWh] 2022 2030 2040 -4 Nuclear For Denmark, the implementaion of Viking Link leads to reduced -5 curtailment of wind for all years. A rise in natural gas and coal -6 generation can be seen in 2022. In 2022 and 2030 biomass production increases as well.

Changein power generationVikingLink -7 -8 2022 2030 2040

36 Electricity prices

Comparison DK_W and GB Danish electricity and wind prices

The inclusion of the Viking Link in the system raises the electricity 450 30% prices in Denmark by approx. 15 DKK/MWh in 2022 while 400 28% reducing average prices in Great Britain. 350 26%

500 160 24% [%] DKK/MWh] 300 450 140 22% 250 400 price [ 20% 120 200 350 18%

100 150 Windprice gap

300 16% Electricity Electricity 250 80 100 14%

200 60 50 12% [DKK/MWh] 150 0 10% 40

100 2010 2020 2030 2040 Electricityprice [DKK/MWh] 20 El-price Base El-price Viking 50 Absolute difference and DK GB Wind price Base Wind price Viking 0 0 Wind price difference Base Wind price difference Viking 2014 2020 2022 2030 2040 Base - DK_W Base - GB VikingLink - DK_W Both the average electricity prices and the wind prices are higher with the Viking Link in operation, the increase in the wind prices is VikingLink - GB Base - Diff VikingLink - Diff more pronounced however, and the percentage difference between electricity and wind price therefore decreases, showing better integration.

37 Price difference DK West and GB duration curve Viking Link scenario

1.500

1.000

500

0

1

231 461 691 921

1611 4141 1151 1381 1841 2071 2301 2531 2761 2991 3221 3451 3681 3911 4371 4601 4831 5061 5291 5521 5751 5981 6211 6441 6671 6901 7131 7361 7591 7821 8051 8281 8511

-500 Electricity price [DKK/MWh] price Electricity

-1.000 2020 2022 2030 2040

Number of Number of Average of Congestion The price differences between Denmark and Great Britain are hours higher hours higher absolute price rent [billion lower when the Viking Link is used. The same trends as in the GB price DK_W price difference DKK] base scenario remain with a decreasing absolute average until 2020 8469 197 148.3 - 2030 and a steep increase in 2040. 2022 8097 505 83.0 0.9 The sum of all hourly price differences shows the annual congestion rent when multiplied by the flow. 2030 6017 2317 76.4 0.8 2040 5090 3397 138.7 1.5

38 Vikinglink import and export

6 4 2 0 -2 2020 2022 2030 2040 -4 -6

Viking link import to DK to link importViking -8 -10 -12 VikingLink - Import VikingLink - Export VikingLink - Net Import

In the figure above, the green represents the annual transmission from Great Britain to Denmark, where orange shows the transmission in the opposite direction. The blue line represents the sum of the two. In 2022 the transmission in the Viking Link is primarily used for export from Denmark to Great Britain as the electricity prices in Great Britain are higher than in Denmark around the year. Towards 2040 the link is used increasingly for transmission in both directions as more and more hours occur where the electricity prices are higher in Denmark than in Great Britain. This is a result of the increased volatility of the hourly electricity prices and the different generation profiles for wind power in Great Britain and Denmark.

39 Model results HEAT PUMPS SCENARIO

40 Heat pump capacity

Heat pumps scenario In the heat pump scenario, the heat pump capacity is found by reducing the total tax and tariff for electricity usage for district heating generation Installed heat Capacity in natural FLH to 150 DKK/MWh and letting the model optimize investments in the capacity [MW] gas areas power and district heating sector in Denmark. The level of the chosen Heatpumps tariff is meant to reflect the socio-economic tariff of using electricity. 2020 369 96% 6,242 However, no detailed analysis on this matter has been carried out in this 2022 490 97% 5,769 project, and 150 DKK/MWh is a bit lower, than the socio-economic tariff of around 160 DKK/MWh (excluding losses) the Danish Energy Agency is 2030 2,299 40% 3,313 using for industrial consumers. However, heat pumps can have 2040 3,976 24% 3,271 production patterns better suited for the grid capacity compared to Electric boilers industrial consumers, which might be able to justify even lower socio- 2020 149 68% 1,210 economic tariffs. 2022 133 66% 732 In order to ensure a reasonable pace of introduction of heat pumps and 2030 281 11% 512 electric boilers, the maximum allowed capacity in 2020 and 2022 is 2040 251 0% 752 capped at 300 and 500 MW heat capacity, compared to the capacity in 2014. The total heat pump and boiler capacities in the Heatpumps 5000 scenario are given in the left table, along with their resulting full load hours. 4000 Technical data for the heat pumps are based on the Technology

&boiler 3000 catalogue by the Danish Energy Agency and Energinet.dk. 2000 Investment cost O&M variable O&M Exogenous

1000 COP [mio. DKK/MWh [DKK/MWh heat [DKK/MW heat Heatpump capacity DK in [MW] heat] gen.] cap.] 0 2020 2014 2020 2022 2030 2040 2.9 5.0 2.4 14,500 Year 2014 Base Heatpumps 2030 3.0 4.6 2.4 14,500

41 Generation

Heat generation

160 Solar Initially, mainly heat generation by natural gas and solar is 140 replaced by the new heat pumps. From 2030 onwards heat Electric 120 generation from biomass is replaced. Bio-oil 100 50 80 Biogas 60 Wood pellets 40

Heatgeneration [PJ] 40 Straw 30

20 [PJ] Solar Wood 20 0 Electric Oil 10

Wood pellets

Base Base Base Base Natural gas 0 Straw

Coal -10 Wood

Heatpumps Heatpumps Heatpumps Heatpumps 2020 2022 2030 2040 Waste -20 Natural gas Waste

Surplus heat -30 Changein heat generation -40

The heat generation in Denmark shows a gradual increase in heat -50 generation by heat pumps in the Heatpumps scenario. In 2020 2020 2022 2030 2040 the heat generation from boilers and heat pumps is only 6.5% of the total heat generation. By 2040 36% of heat generation is from electricity.

42 Electricity prices

Comparison DK_W and GB Danish electricity and wind prices

The Heat pumps scenario raises the electricity prices in 500 30% Denmark in the longer term with approx. 10-20 DKK/MWh by 28% 2030 and 2040, while prices in Great Britain are largely 400 26% unaffected. All else equal, the decreasing gap could slightly

24% [%] decrease the potential for connecting the two regions. DKK/MWh] 300 22% 500 160 price price [ 20% 450 140 200 18% 400 120 16% Windprice gap 350 Electricity 100 14% 300 100 12% 250 80 0 10%

200 60 2010 2020 2030 2040 [DKK/MWh] 150 El-price Base El-price Heatpumps 40 Wind price Base Wind price Heatpumps 100 Electricityprice [DKK/MWh] Wind price difference Base Wind price difference Heatpumps 20 50 Absolute difference and DK GB Both electricity and the wind prices are higher in the Heat 0 0 pumps scenario, the increase in the wind prices is more 2014 2020 2022 2030 2040 pronounced however, and the percentage difference Base - DK_W Base - GB Heatpumps - DK_W between electricity and wind price therefore decreases Heatpumps - GB Base - Diff Heatpumps - Diff compared to the base scenario.

43 Model results FLEXIBILITY ON LARGER POWER PLANTS SCENARIO

44 Capacity

Added boiler capacity in Denmark The boilers that were added to the central areas in Denmark have the following characteristics 4.000

3.500 O&M variable Investment Efficiency [DKK/MWh heat gen.] [DKK/MW heat cap.] 3.000 99% 4.0 0,5

2.500 capacity[MW] 2.000 The total heat capacity and the FLH of the boiler are shown for the 1.500 four years in the table below. In 2040 about 400 MW was already

boiler boiler heat in the system in the base scenario and is added to the additional 1.000 capacity. The low number of full load hours indicates the limited 500 usage of the electric boilers.

Additional Total installed heat 0 FLH 2020 2022 2030 2040 capacity [MW] 2020 3,591 269 On the larger power plants in Denmark, additional electric boiler capacity is installed to ensure more flexibility. To some extent, a 2022 3,753 262 similar measure can be achieved by adding the option of turbine 2030 2,989 444 bypass to the power plants. Therefore, no taxes and tariffs on 2040 2,431 818 electricity usage by the electric boilers are assumed. However, it should be noted that the electric boilers can also run when the power plants are not in operation and thus do not only act like turbine bypass.

45 Change in generation in Denmark

Electricity generation Heat generation The added boilers generate heat that replaces mainly biomass, 600 natural gas and coal in the system. The displaced biomass is mainly CHP production. 400

Biomass 8000 200 Biogas Other fossils 6000 Wind Biomass 0 ]

Electric [TJ 4000 Biogas

-200 Solar Electric 2000 Coal Solar -400 Waste 0 Surplus heat Natural gas

Change in power in Change power [GWh] generation -600 -2000 Bio-oil

Coal -800 -4000

2020 2022 2030 2040 Changein heat generation Waste -6000 Natural gas

In Denmark, the changes in the electricity production show a clear -8000 trend to less curtailment and a reduction of power generation 2020 2022 2030 2040 from biomass CHP. In the earlier years some changes in natural gas and coal are also seen.

46 Change in overall electricity generation

Overall electricity generation

The changes in the electricity generation in the entire model for the increased flexibility on central power plants are for a large part a reduction in curtailment of wind power.

47 Boiler generation and electricity price

Boiler generation (2020) Boiler electricity price (2020)

1000 3500 1000

800 3000 800

2500 600 600 2000 400 400 1500 200 200

1000

Electricityprice [DDK/MWh] Electricityprice [DDK/MWh]

0 500 0

1 1

438 875 399 797

1312 1749 2186 2623 3060 3497 3934 4371 4808 5245 5682 6119 6556 6993 7430 7867 8304 1195 1593 1991 2389 2787 3185 3583 3981 4379 4777 5175 5573 5971 6369 6767 7165 7563 7961 8359 -200 0 -200

Avg. el-price Boiler generation Avg. el-price Boiler el-price

The electric boilers generate more heat at low electricity prices, The graph above shows the electricity price duration curve and while heat generation is based on CHP at higher electricity prices. the duration curve of the electricity price when the electric boilers Compared to the base scenario, the number of hours with very are in operation. It can be seen that the boilers are only in low or negative electricity prices in Denmark in 2020 is reduced operation for a maximum of 2000 hours a year at low electricity significantly, which leads to a relative large impact on the prices. The graph does not indicate whether boilers run at full electricity prices captured by wind power. capacity all over Denmark, and thus the 2000 hours are not equivalent to the number of full load hours.

48 Electricity prices

Comparison DK_W and GB Danish electricity and wind prices The flexibility on the central power plants in Denmark raises the 450 30% electricity prices in Denmark, while prices in Great Britain are 400 28%

unaffected. The gap between the two countries is therefore [%] 26% lowered in the CPP flex scenario. 350 24%

DKK/MWh] 300 22% 250 500 160

price price [ 20% 200 450 140 18% Windprice gap 150 400 16% 120 350 Electricity 100 14% 300 100 50 12% 250 80 0 10% 2010 2020 2030 2040 200 60 150 [DKK/MWh] El-price Base El-price CPP_flex 40 Wind price Base Wind price CPP_flex 100

Electricityprice [DKK/MWh] Wind price difference Base Wind price difference CPP_flex 20 50 Absolute difference and DK GB Both electricity and the wind prices are higher in the CPP flex 0 0 scenario compared to the Base case. The increase in the wind 2014 2020 2022 2030 2040 Base - DK_W Base - GB CPP_flex - DK_W prices is more pronounced, especially in the early years and the percentage difference between electricity and wind price CPP_flex - GB Base - Diff CPP_flex - Diff therefore decreases compared to the base scenario.

49 Model results INDIVIDUAL HEAT PUMPS SCENARIO

50 Capacity

Added storage capacity in Denmark

1400 The graph on the left illustrates the amount of electricity, that can be “stored” by using the heat pumps in a more 1200 flexible way. During summer, the lower heat demand 1000 limits the option to “store” electricity, which is taken into 800 account by using a seasonal variation of the maximum storage capacity, illustrated on the graph below. 600

400 1 Storage Storage capacity[MWh] 200 0,9 0,8 0 0,7 2020 2022 2030 2040 0,6 West Denmark East Denmark 0,5 0,4

In the Individual Heat pumps scenario, electricity usage for Derate ratio 0,3 individual heat pumps has increased flexibility. While the electricity demand for individual heat pumps follows the heat demand in the 0,2 base scenario, the model has the option to shift demand around 4 0,1

hours in time in the individual heat pump scenario. This represents 0

1 761

the option to use local hot water tanks and the heat stored in 381

1521 1901 2281 2661 3041 3421 3801 4181 4561 4941 5321 5701 6081 6461 6841 7221 7601 7981 8361 buildings to use the heat pumps at times with lower power prices. 1141 West Denmark East Denmark

51 Change in generation in Denmark

Electricity generation

25

20

15 Biomass 10 Biogas 5 Wind Solar 0 Coal -5 Waste -10 Natural gas

Change in power in Change power [GWh] generation -15

-20 2020 2022 2030 2040

The changes in the electricity generation in Denmark for the Individual heat pumps scenario are limited. In 2022 and 2030 small some reduction in wind curtailment is seen. The limited impact on the overall power system also limits the impact on wind power integration and system economy.

52 Electricity prices

Comparison DK_W and GB Danish electricity and wind prices Negligible changes compared to base scenario 500 30% 28% 500 160 400 26% [%] 450 140 24%

400 DKK/MWh] 300 22% 120 350 20%

300 100 price [ 200 18% Windprice gap 250 80 16% 100 14% 200 60 Electricity 12%

150 [DKK/MWh] 40 0 10% 100 2010 2020 2030 2040

Electricityprice [DKK/MWh] 50 20 El-price Base

Absolute difference and DK GB El-price Indiv_heatpumps 0 0 Wind price Base 2014 2020 2022 2030 2040 Wind price Indiv_heatpumps Wind price difference Base Base - DK_W Base - GB Wind price difference Indiv_heatpumps Indiv_heatpumps - DK_W Indiv_heatpumps - GB Base - Diff Indiv_heatpumps - Diff Negligible changes compared to base scenario

53 Model results INTEGRATION INDICATORS

54 Additional power demand, export and flexibility

To some extent, the effect of the different scenarios on the overall power system and the integration of wind power is limited by the size of the integration measure. The different scenarios do not necessarily compare in terms of size of the integration measure with respect to power capacity, power demand or associated investment costs. The table below gives an overview on the size of the different integration measures from a power system point of view, when compared to the base scenario. Note, that the shown power “demand” is not necessarily additional demand, but an interpretation of the effects for the Danish power system. As an example, export from Denmark through Viking Link is shown as additional demand, even though the total demand in the overall system is unchanged. Please see the notes for further details. The values for power demand are based on the actual model results and not an a priori assumption.

Additional potential power ”demand” [GW] VikingLink Heatpumps CPP flex Indiv heatpumps Combi 1 Combi 2 Combi 3 2020 0.00 0.09 3.65 0.15 0.09 3.65 3.74 2022 1.40 0.13 3.81 0.18 1.53 5.21 5.34 2030 1.40 0.49 3.04 0.34 1.89 4.44 4.93 2040 1.40 1.00 2.04 0.56 2.40 3.44 4.44 Additional power ”demand" [TWh] VikingLink* Heatpumps CPP flex Indiv heatpumps** Combi 1*** Combi 2**** Combi 3***** 2020 0.00 0.62 0.97 0.42 0.62 0.97 1.52 2022 9.78 0.81 0.99 0.53 10.49 10.34 10.99 2030 4.40 1.58 1.34 0.99 5.43 5.44 6.29 2040 1.89 3.30 1.76 1.64 4.45 2.96 4.98

*Power demand shows net export from Denmark to Great Britain on an annual basis **Power demand shows amount of power demand, that is made flexible. Annual power demand does not change. ***Power demand shows combination of net export from Denmark to Great Britain on an annual basis and additional annual power demand for heat pumps. **** Power demand shows combination of net export from Denmark to Great Britain on an annual basis and additional annual power demand for electric boilers. *****Power demand shows combination of net export from Denmark to Great Britain on an annual basis and additioanl power demand from heatpumps and electric boilers

55 Wind weighted prices in Denmark

Relative difference Ratio of differences 30% To validate the wind weighted prices of the integration scenarios 25% compared to the base scenario, the following measure can be used

[%] 20% ∆ 푊𝑖푛푑 푤푒𝑖𝑔ℎ푡푒푑 푝푟𝑖푐푒 ∆ 퐴푣푒푟푎𝑔푒 푒푙푒푐푡푟𝑖푐𝑖푡푦 푝푟𝑖푐푒 15% The table below shows the changes of the wind weighted and average electricity price as well as the ratio. A high ratio indicates good wind 10%

integration. The ratio is generally higher for the Viking Link scenario Windprice gap 5% compared to the Heat pumps scenario. The CPP_flex scenario shows the highest values for 2020 and 2022. 0% The combi scenarios show changes in wind weighted and average 2010 2020 2030 2040 electricity prices, that are close to the sum of the respective individual Base VikingLink Heatpumps scenarios. CPP_flex Indiv_heatpumps Combi 1 Fraction (diff. Wind weighted price/diff average electricity price) Combi 2 Combi 3 Indiv.Heat- VikingLink Heatpumps CPP flex Combi 1 Combi 2 Combi 3 pumps The wind weighted price is an indicator for wind integration; in a system 2020 0/0 4.5/2.9 29.9/13.1 0.2/0 4.5/2.9 29.9/13.1 32.6/15.3 with good wind integration the wind weighted price does not differ 2022 23.2/14.7 3.6/3 26.6/13.8 0.3/0.2 28.2/17.8 44.6/25.1 47.2/26.9 largely from the average electricity price. All analysed integration 2030 15.9/8.6 13.5/9.7 15.8/10.8 0.2/-0.1 29/17.6 34.5/19.5 47.5/27.5 measures show a positive effect on wind integration in terms of the wind 2040 16.5/8.7 32.4/21.1 21.6/13.2 0.3/0.1 60.1/35.4 38.9/21 71.9/41.2 weighted electricity price. (the Individual heat pumps scenario is shown Ratio Indiv.Heat- in dotted lines as the effect is small). In the short term, the effect is most VikingLink Heatpumps CPP flex Combi 1 Combi 2 Combi 3 pronounced in the CPP_flex scenario. However, also the introduction of pumps 2020 - 1.52 2.29 - 1.52 2.29 2.13 Viking Link in 2022 leads to a decreasing gap compared to 2020. In early 2022 1.58 1.21 1.93 1.52 1.59 1.78 1.75 years, the heat pumps scenario shows very low differences, but by 2040 2030 1.84 1.40 1.46 -1.59 1.64 1.77 1.73 the values resemble those of the other scenarios. In general, onshore 2040 1.90 1.53 1.64 5.75 1.70 1.85 1.75 wind has a larger price gap than offshore wind.

56 Overall curtailment

Wind curtailment RE curtailment

1,80% Similar results are seen for the overall RE curtailment (which also includes wind, solar and hydro run-of-river curtailment). 1,60% 1,40% 1,20% 1,40% 1,00% 1,20% 0,80% 0,60% 1,00%

0,40% 0,80% 0,20%

Relative Relative windcurtailment [%] 0,60% 0,00% 2010 2015 2020 2025 2030 2035 2040 2045 0,40% Base VikingLink Heatpumps 0,20%

CPP_flex Indiv_heatpumps Combi 1 curtailment Relative RE [%] Combi 2 Combi 3 0,00% Wind curtailment is another indicator to assess wind integration. A 2010 2015 2020 2025 2030 2035 2040 2045 drop in overall wind power curtailment is seen for the Viking Link and Base VikingLink Heatpumps Combi 1 scenario in 2022. The Heat pumps scenario has a smaller CPP_flex Indiv_heatpumps Combi 1 effect in 2020, but is similar in curtailment level to the Viking Link Combi 2 Combi 3 scenario by 2030. The CPP flex scenario shows the largest effect in all years but 2040.

57 Curtailment in Denmark

Wind curtailment

3,50% A drop in Danish wind power curtailment is seen for the Viking Link and Combi 1 scenario in 2022. The Combi 1 scenario shows a large reduction 3,00% in curtailment compared to the single integration scenarios. For the scenario with flexibility on Danish power plants, very little curtailment is 2,50% seen in 2020 and 2022. In later years this scenario shows curtailment in the order of the Combi 1 case. Overall the difference from the base 2,00% scenario is small. Both combination scenario 2 and 3 show low curtailment for all years. 1,50% Geographical distribution of curtailment will in reality to some extent 1,00% depend on the detailed regulation regimes in different countries, which are not replicated in the model.

0,50% Relative Relative windcurtailment [%] 0,00% 2010 2015 2020 2025 2030 2035 2040 2045 Base VikingLink Heatpumps CPP_flex Indiv_heatpumps Combi 1 Combi 2 Combi 3

58 Model results ECONOMY OF WIND INTEGRATION MEASURES

59 • Evaluation of economic effects based on – Comparison to Base scenario – Socioeconomic perspective: Changes in system costs for providing heat and electricity – All numbers rounded to multiple of 50 million DKK Economy • Included elements – Capital cost on technologies. Based on a socioeconomic interest rate of 4% and a Socio-economic lifetime of 20 years. Including changes in capital cost due to the introduction of heat perspective pumps and replacement of CHP capacity. – Fixed and variable O&M – Fuel cost

– CO2-damage cost (value of CO2 set equal to ETS price)

• No socio-economic value of moving emission from the UK (Higher tax on CO2 before 2030) to continental Europe (lower price on CO2 before 2030)

– Value of changes in hydro generation. Recalculated to fuel, O&M and CO2 savings/cost based on the average emission factor for electricity generation in the overall system excluding hydro, wind, solar and municipal solid waste, a general O&M cost of 30 DKK/MWh and the remaining on fuel cost to match the overall value of changes in hydro generation. – For the NPV calculation, the period from 2020 to 2051 is taken into account (30 years lifetime on Viking Link). Since no model calculation has been performed beyond 2040, the same values as for 2040 are assumed for the period 2040-2051. • Not included elements – Cost of establishing flexibility on individual heat pumps

– Damage costs of changes in other emissions (NOx and SO2). This could have a (smaller) impact on the results. However, it is difficult to estimate the emission factors for the existing power plants and a detailed analysis of this matter was not carried out in this project. – Cost due to tax distortion

60 Viking Link scenario Economy Annual savings

• Economy includes 13.4 billion DKK investment costs for the Viking 2.000 Link (Source: ENTSO-E Ten Year Network Development Plan) and an estimated 0,5% of the investment for Operation and ) 1.500 maintenance/year. • Main savings from reduced fuel cost 1.000 900 • In the short run, annual operational savings are below the capital and O&M cost for Viking Link and the socio-economic value for 500

2022 alone is therefore negative. million DKK/year 50 – Price difference between Denmark and Great Britain is - - -200 partly driven by different CO2-taxes, which do not reflect socioeconomic savings. -500 • Savings increase, even though average price difference between -1.000 Denmark and Great Britain decreases Annualsavings ( – Price spread by hour increases -1.500 – In the long run, additional transmissions capacity can 2020 2022 2030 2040 reduce wind curtailment • Socioeconomicy is shown for the entire region. The effect on Capital Cost (mDKK) Fuel Cost (mDKK) individual countries will differ. For Denmark the socioeconomic O&M Cost (mDKK) CO2 Damage Cost (mDKK) value of Viking Link is positive throughout the entire period, while Total other countries experience losses in the short run. • By 2040, the general value of new transmission in the system • NPV in 2020: approx 5.6 billion DKK (30 yrs.) increases due to higher shares of variable RE generation. Viking Link is the only allowed transmission investment over the period. However, other transmission lines could also show positive economy.

• Higher CO2-emissions in the short run due to lower generation from gas in the UK and higher generation from coal in continental Europe

61 Heat pumps scenario

Economy Annual savings

• Economy calculations include capital cost and savings in the 4.000 Danish district heating system. 3.500 • Main savings from reduced fuel cost, but also capital cost. 3.000 • Positive savings, from first year, although low for 2020 and 2.500 2.200 2022. Increasing value as a result of higher usage of heat 2.000 pumps. 1.500 • Higher CO -emissions due to higher electricity consumption 2 1.000 950 and reduced electricity generation from biomass in Denmark. 500 In a well-functioning CO quota system, one could argue that 2 - - 50 the short term effect would be rising CO2-prices and no -500 additional CO2-emissions, possibly with a similar economic effect. Annualsavings (million. DKK/year) -1.000 • The model is not allowed to invest in new capacity in other -1.500 power production when the heat pumps are introduced and 2020 2022 2030 2040 the increased need for electricity therefore comes from Capital Cost (mDKK) Fuel Cost (mDKK) existing thermal plants. A different approach where the O&M Cost (mDKK) CO2 Damage Cost (mDKK) model would be allowed to invest in wind power could Total change the results on CO2-emissions.

• NPV in 2020: approx. 20.2 billion DKK

62 Flexibility on Larger Power Plants Scenario

Economy Annual savings

• Economy calculations include changes in capital cost but not 700 fixed O&M costs connected to introducing flexibility on the 600 power plants. 500 500 • Main savings arise from reduced fuel cost. 400 • Increasing savings as a result of higher number of hours with low electricity prices and subsequently higher number of full 300 load hours for electric boilers. 200 250 100 50 100 - -100

Annual savings Annualsavings (billionDKK/year) -200 -300 2020 2022 2030 2040

Capital Cost (mDKK) Fuel Cost (mDKK) O&M Cost (mDKK) CO2 Damage Cost (mDKK) Total

• NPV approx. 5.1 billion DKK

63 Individual heat pumps scenario

Economy Annual savings

• Economy calculations do not investment cost arising from 250 flexibilisation of individual heat pumps, such as individual heat storages. This cost has not been estimated within the 200 current project. 150 150 • Main savings from reduced fuel cost. DKK/year)

• Small savings in the short run. 100 (million (million 50 50 - - -

-50 Annual savings Annualsavings

-100 2020 2022 2030 2040

Capital Cost (mDKK) Fuel Cost (mDKK) O&M Cost (mDKK) CO2 Damage Cost (mDKK) Total

• NPV approx. 1.3 billion DKK

64 Combined scenario 1 : Heatpumps + Viking Link

Economy Annual savings

• Economy calculations include capital cost and savings in the 4.000 Danish district heating system, and capital costs and O&M costs for the Viking Link. 3.000 2.850 • Main savings from reduced fuel cost • Low savings in the short run, in parts due to the low savings in 2.000 the heat pump scenario, and in parts due to the extra cost 1.000 from the Viking Link scenario for 2022. 800 • Increasing value towards 2030 and 2040 as a result of higher - - fluctuations in electricity prices and higher usage of heat -150 pumps. -1.000 • Higher CO2-emissions due to higher electricity consumption Annualsavings (mio. DKK/year) and reduced electricity generation from biomass in Denmark -2.000 2020 2022 2030 2040

Capital Cost (mDKK) Fuel Cost (mDKK) O&M Cost (mDKK) CO2 Damage Cost (mDKK) Total

• NPV approx. 22.8 billion DKK • NPV for Viking Link with heat pump scenario as base: 2.6 billion DKK corresponding to approx 47% of the value of Viking Link compared to the base scenario

65 Combined scenario 2: CPP flex + Viking Link

Economy Annual savings

• Economy calculations include changes in capital cost but not 2.500 fixed O&M costs connected to introducing flexibility on the power plants. Viking Link investment costs and O&M costs 2.000 are included. 1.500 • Main savings from reduced fuel cost 1.000 1.100 • Low savings in the short term. Increasing value as a result of higher fluctuations in electricity prices and increased usage of 500 electric boilers. 150 - 50 -200 -500

-1.000 Annual savings Annualsavings (million. DKK/year)

-1.500 2020 2022 2030 2040

Capital Cost (mDKK) Fuel Cost (mDKK) O&M Cost (mDKK) CO2 Damage Cost (mDKK) Total

• NPV approx. 7.6 billion DKK • NPV for Viking Link with CPP flex scenario as base: 2.4 billion DKK corresponding to approx 43% of the value of Viking Link compared to the base scenario

66 Combined scenario 3: Heatpumps + CPP flex + Viking Link

Economy Annual savings

• Economy calculations include capital cost and savings in the 5.000 Danish district heating system including changes in capital cost but not fixed O&M costs connected to introducing 4.000 flexibility on the power plants. Viking Link investments and O&M costs are included. 3.000 3.150 • Main savings from reduced fuel cost, but also capital cost 2.000 • Low savings in the short term. Increasing value as a result of higher fluctuations in electricity prices and higher usage of 1.000 950 heat pumps. - 200 • Higher CO2-emissions due to higher electricity consumption -300 and reduced electricity generation from biomass in Denmark

Annual savings Annualsavings (mio. DKK/year) -1.000

-2.000 2020 2022 2030 2040

Capital Cost (mDKK) Fuel Cost (mDKK) O&M Cost (mDKK) CO2 Damage Cost (mDKK) Total

• NPV approx. 25.3 billion DKK • NPV for Viking Link with heat pump + CPP flex as base: 1.3 billion DKK corresponding to approx 23% of the value of Viking Link compared to the base scenario

67 Socioeconomic value of different scenarios

The net present value (NPV) of the different scenarios is calculated for 3500 2020 as base year with a discount rate of 4% and over the period from 3000 2020 to 2051, which corresponds to a lifetime of Viking Link of 30 years. All scenarios show a positive socioeconomic net present value. However, 2500 the individual heatpump scenario does not include possible investment 2000 cost, which could change this picture. For all scenarios, savings are largest in the long term for 2030 and beyond, while savings are low or negative in 1500 the short term. 1000 When comparing the combination scenarios with their respective scenarios without the Viking Link, it is possible to estimate the value of 500 Viking Link with a different base assumption. This comparison show, that the NPV of Viking Link is reduced to between 27% and 47% when all other 0 Annual savings Annualsavings [million DKK/year] 2020 2022 2030 2040 integration measures would be in place, compared to a situation, where -500 only Viking Link is included. However, this shows the total NPV including the investment and O&M cost for the VikingLink. When only looking at the VikingLink Heatpumps CPP_flex Indiv_heatpumps operational savings in the system, the value of Viking Link is reduced to Combi 1 Combi 2 Combi 3 between 78% and 84%. This means, that a large share of the operational savings from Viking Link can still be achieved, even if all other integration measures are implemented. Scenario NPV NPV of VikingLink [billion NPV of Viking Link compared NPV of operational savings [billion DKK] DKK] to Viking Link without other from Viking Link compared to integration measures Viking Link without other integration measures VikingLink 5.6 5.6 100% 100%

Heatpumps 20.2 -

CPP_flex 5.1 -

Indiv_heatpumps 1.3 - Combi 1 22.8 2.6 47% 84% (Heatpumps+VikingLink) Combi 2 7.6 2.6 45% 84% (CPP_flex+VikingLink) Combi 3 68 25.3 1.3 27% 78% (Heatpumps+CPP_flex+VikingLink)