Wissenschaftliche Aussprache zur Dissertation Integrating variable electricity supply
Wissenschaftliche Aussprache zur Dissertation Integrating variable electricity supply from wind and solar PV into power systems vorgelegt von Diplom-Physiker Falko Ueckerdt Fakultät VI – Planen Bauen Umwelt der Technischen Universität Berlin zur Erlangung des akademischen Grades Doktor der Naturwissenschaften Dr. rer. nat. Berlin, den 15. Juli 2014
Global electricity generation 2013 <4 % wind & solar My thesis aims at understanding variability to improve the economic evaluation of wind and solar PV ? ~70 % Fossil fuels Data from REN 21 (2014) 2
What are redistribution effects due to VRE support policies? (paper 6) What is the link between the marginal value and integration costs literature? (paper 4) How do integration costs affect the optimal share of variable renewables? (paper 3) What are the full costs of variable renewables? (paper 2) How to improve the representation of variability in integrated assessment models? (paper 5) What are major integration challenges of variable renewables? (paper 1) 3
What are redistribution Hirth effects due&to. Ueckerdt VRE support policies? Energy Policy 2013 (paper 6) Müller, the What Ueckerdt, is the link between Hirth, Nicolosi marginal value and integration costs literature? submitted to (paper Energy 4) Renewable Hirth, How Ueckerdt, do integration Edenhofer costs affect the optimal share of variable submitted renewables? to Renewable (paper. Energy 3) Ueckerdt, Hirth, What are the full costs of Luderer, Edenhofer variable renewables? (paper 2) Energy 2013 Ueckerdt, How to. Brecha, improve. Luderer, the Sullivan, representation Schmid, of. Bauer, variability Böttger in integrated assessment models? submitted (paper to 5) Energy What Ueckerdt, are major Brecha, integration Luderer challenges of variable submitted renewables? to Renewable (paper. Energy 1) 4
What are redistribution effects due to VRE support policies? (paper 6) What is the link between the marginal value and integration costs literature? (paper 4) How do integration costs affect the optimal share of variable renewables? (paper 3) What are the full costs of variable renewables? (paper 2) How to improve the representation of variability in integrated assessment models? (paper 5) What are major integration challenges of variable renewables? (paper 1) 5
What are redistribution effects due to VRE support policies? (paper 6) What is the link between the marginal value and integration costs literature? (paper 4) How do integration costs affect the optimal share of variable renewables? (paper 3) What are the full costs of variable renewables? (paper 2) How to improve the representation of variability in integrated assessment models? (paper 5) What are major integration challenges of variable renewables? (paper 1) 6
Variable renewables (VRE) have three characteristic properties Output is variable in time Output is uncertain Output is variable in space Forecast MW MW Load Wind power 1 year matching of demand supply with wind is costly „Profile costs“ Wind power 1 day short-term balancing of supply and demand is costly „Balancing costs“ Integration costs transmission is costly „Grid costs“ 7
The Residual Load Duration Curve (RLDC) Load curve (demand) German data MW 1 year t Ueckerdt, Brecha, Luderer. „What are major integration challenges? “ 8
The Residual Load Duration Curve (RLDC) Load Duration Curve Load curve (demand) German data MW MW Sorting 1 year t 1 year (sorted) Ueckerdt, Brecha, Luderer. „What are major integration challenges? “ 9
The Residual Load Duration Curve (RLDC) Load Duration Curve Load curve (demand) German data MW 1 year MW t 1 year (sorted) Load and renewable supply MW 1 year t Wind Ueckerdt, Brecha, Luderer. „What are major integration challenges? “ 10
The Residual Load Duration Curve (RLDC) Load Duration Curve Load curve (demand) German data MW 1 year MW t 1 year (sorted) Load and renewable supply MW Residual load curve MW Load – renewables 1 year t t Wind 1 year Ueckerdt, Brecha, Luderer. „What are major integration challenges? “ 11
The Residual Load Duration Curve (RLDC) Load Duration Curve Load curve (demand) German data 1 year MW t Load and renewable supply MW Sorting MW 1 year (sorted) Residual load curve MW 1 year t t Wind 1 year Ueckerdt, Brecha, Luderer. „What are major integration challenges? “ 12
The Residual Load Duration Curve (RLDC) Residual Load Duration Curve Load curve (demand) German data 1 year MW t Load and renewable supply MW Sorting MW MW 1 year (sorted) Residual load curve t t 1 year Ueckerdt, Brecha, Luderer. „What are major integration challenges? “ 13
The Residual Load Duration Curve (RLDC) Residual Load Duration Curve Load curve (demand) German data MW MW Wind 1 year t 1 year (sorted) Load and renewable supply MW MW 1 year Residual load curve t t Wind 1 year Ueckerdt, Brecha, Luderer. „What are major integration challenges? “ 14
higher the challenges increase RLDCWith capture thepenetrations major impacts of temporal variability 65% share 1 0 65% 35% 50% 12% share Load (normalized to peak load) 1 year (sorted) Overproduction (56%) Wind : Solar 1: 3 Low 1 capacity credit 0 German data (2011) 1 year 1(sorted) year (sorted) Reduced full-load hours Overproduction (1%) (5%) Overproduction (17%) Wind : Solar 3: 1 These three challenges induce profile costs We measure those challenges in an extensive data analysis Ueckerdt, Brecha, Luderer. „What are major integration challenges? “ 15
Integration challenges depend on penetration, VRE mix and regional patterns of VRE and load Challenge parameter 1: capacity credit US Midwest Win d pe netr atio n Germany r Sola ion pe at netr Win d pe netr atio n r Sola paper 1: Ueckerdt, Brecha, Luderer. „Major integration challenges? “ pe ion at netr 16
Integration challenges depend on penetration, VRE mix and regional patterns of VRE and load Challenge parameter 1: capacity credit US Midwest Germany paper 1: Ueckerdt, Brecha, Luderer. „Major integration challenges? “ 17
What are redistribution effects due to VRE support policies? (paper 6) What is the link between the marginal value and integration costs literature? (paper 4) How do integration costs affect the optimal share of variable renewables? (paper 3) What are the full costs of variable renewables? (paper 2) How to improve the representation of variability in integrated assessment models? (paper 5) What are major integration challenges of variable renewables? (paper 1) 18
€/MWh What are the full costs of variable renewables? Wind LCOE Coal LCOE 19 IEA & NEA 2011, EIA 2013, IRENA 2012, IPCC 2011, Nitsch et al. 2011, Kost et al. 2012, Karlynn & Schwabe 2009,
What are the full costs of variable renewables? • The new metric allows comparing technologies and should replace LCOE • integration costs are derived from a qualitative definition: €/MWh “all additional costs in the non-VRE part of the power system when adding VRE” • A quantification of System LCOE requires the decomposition Wind Integration Wind LCOE Costs System LCOE Ueckerdt, Hirth, Luderer, Edenhofer. „Full costs of variable renewables? “ 20
What are the full costs of variable renewables? €/MWh • Profile costs were estimated from an own model for Germany profile balancing costs • Grid and balancing costs from a literature review grid costs • Limitations: only power sector considered, no import/export, no demand response, no storage Integration costs Wind Integration LCOE Costs Wind System LCOE Ueckerdt, Hirth, Luderer, Edenhofer. „Full costs of variable renewables? “ 21
Variability matters: high integration costs at high VRE shares System LCOE of wind Changing residual capacity Grid costs Balancing costs Integration costs Profile costs Ueckerdt, Hirth, Luderer, Edenhofer. „Full costs of variable renewables? “ 22
What are redistribution effects due to VRE support policies? (paper 6) What is the link between the marginal value and integration costs literature? (paper 4) How do integration costs affect the optimal share of variable renewables? (paper 3) What are the full costs of variable renewables? (paper 2) How to improve the representation of variability in integrated assessment models? (paper 5) What are major integration challenges of variable renewables? (paper 1) 23
How do integration costs affect the optimal share of VRE? Cost perspective Wind System LCOE Integration costs Grid parity €/MWh Wind LCOE Wind Integration Wind LCOE Costs System LCOE Average electricity price q optimal share wind share Hirth, Ueckerdt, Edenhofer. „Economic framework: optimal share? “ 24
Two perspectives on integration costs Value perspective €/MWh Cost perspective Wind Integration Wind LCOE Costs System LCOE Average Integration Wind electricity Costs marginal price value Hirth, Ueckerdt, Edenhofer. „Economic framework: optimal share? “ 25
Two perspectives on integration costs Value perspective €/MWh Wind LCOE Average electricity price Integration costs Wind marginal value q optimal share wind share Hirth, Ueckerdt, Edenhofer. „Economic framework: optimal share? “ 26
Two perspectives on integration costs Cost perspective Value perspective Wind System LCOE Wind LCOE Average electricity price €/MWh Wind LCOE Average electricity price Wind marginal value q q optimal share wind share Hirth, Ueckerdt, Edenhofer. „Economic framework: optimal share? “ 27
What are redistribution effects due to VRE support policies? (paper 6) What is the link between the marginal value and integration costs literature? (paper 4) How do integration costs affect the optimal share of variable renewables? (paper 3) What are the full costs of variable renewables? (paper 2) How to improve the representation of variability in integrated assessment models? (paper 5) What are major integration challenges of variable renewables? (paper 1) 28
What are redistribution effects due to VRE support policies? (paper 6) What is the link between the marginal value and integration costs literature? (paper 4) How do integration costs affect the optimal share of variable renewables? (paper 3) What are the full costs of variable renewables? (paper 2) How to improve the representation of variability in integrated assessment models? (paper 5) What are major integration challenges of variable renewables? (paper 1) 29
Representing residual load duration curves in IAMs • The RLDC endogenously changes controlled by 4 parameters Load (GW) • Changes depend on penetration and mix of VRE • 2 -dimensional functions are derived from data • Power-to-gas storage Curtailment rate Intermediate load triangle Base load box Time (sorted) 1 year Win d pe netr atio n e rp Sola ion at r t e n 30 Ueckerdt, Brecha, Luderer, Sullivan, Schmid, Bauer, Böttger. “Representing variability in IAMs? ”
RLDC approach: variability reduces the power generation from VRE REMIND-D model GHG mitigation scenario for Germany (-80% GHG emissions in 2050 rel. to 1990) 31 Ueckerdt, Brecha, Luderer, Sullivan, Schmid, Bauer, Böttger. “Representing variability in IAMs? ”
RLDC approach: Variability increases mitigation costs Costs of variability +18% Savings due to Power-to-gas -6% REMIND-D model GHG mitigation scenario for Germany (-80% GHG emissions in 2050 rel. to 1990) +12% Additional costs due to RLDC approach 32 Ueckerdt, Brecha, Luderer, Sullivan, Schmid, Bauer, Böttger. “Representing variability in IAMs? ”
The scientific literature was inconclusive about the economic impact of variability Integration costs literature “Estimates of integration costs are low” No rigorous and complete definition of integration costs High detail ? Marginal economic value literature. Integrated assessment models „High impact of variability on marginal values“ „Variable renewables are a prominent mitigation option“ Neglects general equilibrium effects & energy sector links ? Representation of variability needs to be improved Wide scope 33
This thesis bridges the three literature branches Integration costs literature “Estimates of integration costs are low” This thesis 1. Definition of integration costs Marginal value literature. „High impact of variability on marginal values“ 2. Quantification of integration costs 4. Two equivalent perspectives Two methods to represent variability „Variable renewables are a prominent mitigation option“ 1. RLDC approach 2. System LCOE 3. Profile costs No rigorous and complete definition of integration costs This thesis Integrated assessment models Neglects general equilibrium effects & energy sector links Representation of variability needs to be improved 34
Top 4 policy implications 1. A system perspective is required for evaluating VRE. 2. Variability should be considered in market design and policy instruments. 3. System adjustments and integration options can reduce integration costs. 4. High integration costs do not imply that optimal VRE shares are low. Externalities of all technologies need to be internalized. 35
Thank you for your attention!
- Slides: 36