Build wind capacities at windy locations Build wind capacities at windy locations? Assessment of system optimal wind locations under feed-in tariffs Frank Obermüller This presentation: Close the gap between meteorological side (i.e. high resolution wind data) and energy economics (i.e. market value of wind or the impact to the electricity market) 07th January 2017 | Frank Obermüller

Motivation Aim for market integration of renewables IEA Report (June 2016): „[Renewable subsidy] plans should be based on the long-term value of VRE to the power and wider energy system.“ “Next-generation approaches need to factor in the system value of electricity from wind and solar power” Regional wind values | Jan. 07, 2017 | Frank Obermüller

Motivation Value of a wind turbine to the system with:     Market value of wind production: Hirth: “The market value of variable renewables: The effect of solar wind power variability on their relative price." Energy economics 38 (2013) Elberg, Hagspiel: “Spatial dependencies of wind power and interrelations with spot price dynamics.” European Journal of Operational Research 241 (2015) Hirth, Lion. "The market value of variable renewables: The effect of solar wind power variability on their relative price." Energy economics 38 (2013): 218-236. Measures on a national level Finding: Market value of wind is decreasing with increased capacity (Bildchen is nett) Cirtic: No inner-country differentiation with respect to regional wind speeds Elberg, Hagspiel. “Spatial dependencies of wind power and interrelations with spot price dynamics.” European Journal of Operational Research 241 (2015): 260-272. Inner-regional market value of wind is calculated (Focus: Germany) Approach: Copulas – Dependent Distributions of regional wind production time series and the time series of the electricity market price Advantage: Can incorporate simultaneous effects such as cannibalization effects (much capacity at one location reduces market prices) Finding: Regional different market values: Lower at locations with high wind capacity (Bildchen) Critic: No grid congestions are incorporated. In situations with grid congestions, redispatch is necessary and wind in overproduction region has a negative value (since it need to be replaced with more expensive power plants) My approach: Not regional market value of wind but regional system value of wind Incorporated regional market prices, regional wind production, and transmission situation. Value = p_loc * g_loc / p_loc * 1 Extension via high resolution wind data from meteorological institute and wind turbine database Market value Expected market value of wind [EUR/MWh] Regional wind values | Jan. 07, 2017 | Frank Obermüller

Motivation But: Uniform pricing can be inefficient! Market value of wind production: with:     Several electricity markets have uniform pricing, e.g. Europe In uniform pricing, grid situations may not be internalized sufficiently This can lead to inefficient capacity allocations and production situations Improvement: System value of wind production: Hirth, Lion. "The market value of variable renewables: The effect of solar wind power variability on their relative price." Energy economics 38 (2013): 218-236. Measures on a national level Finding: Market value of wind is decreasing with increased capacity (Bildchen is nett) Cirtic: No inner-country differentiation with respect to regional wind speeds Elberg, Hagspiel. “Spatial dependencies of wind power and interrelations with spot price dynamics.” European Journal of Operational Research 241 (2015): 260-272. Inner-regional market value of wind is calculated (Focus: Germany) Approach: Copulas – Dependent Distributions of regional wind production time series and the time series of the electricity market price Advantage: Can incorporate simultaneous effects such as cannibalization effects (much capacity at one location reduces market prices) Finding: Regional different market values: Lower at locations with high wind capacity (Bildchen) Critic: No grid congestions are incorporated. In situations with grid congestions, redispatch is necessary and wind in overproduction region has a negative value (since it need to be replaced with more expensive power plants) My approach: Not regional market value of wind but regional system value of wind Incorporated regional market prices, regional wind production, and transmission situation. Value = p_loc * g_loc / p_loc * 1 Extension via high resolution wind data from meteorological institute and wind turbine database with:     Regional wind values | Jan. 07, 2017 | Frank Obermüller

Methodology Nodal Electricity Market Model Minimization of total system costs Electricity balance restriction (supply=demand) Model and input data are open source Website: www.Pypsa.org Some model facts: Research Focus on Germany Uniform pricing regional different wind situations high share of installed wind capacities 585 nodes 852 lines DC-loadflow model FIAS=Frankfurter Institute for Advanced Science, Renewable Energy Group. Regional wind values | Jan. 07, 2017 | Frank Obermüller

Methodology Nodal Electricity Market Model Additional input: High-resolution wind data FIAS=Frankfurter Institute for Advanced Science, Renewable Energy Group. Regional wind values | Jan. 07, 2017 | Frank Obermüller

Results System value of regional wind production Regional System Value of Wind (Nodal pricing)   Histogram Wind value [100%] Frequency Regional market value of wind (based on zonal pricing, no grid situation included) Value_loc = p*g_loc / p*1 Finding: Differences occur Reason: Different production profiles which are correlated differently as to the market price profile The Regional System Value of Wind Production could have structural breaks, e.g., as in Germany *) Note: Colormap is restricted to a minimum of 80% for comparison reasons Regional wind values | Jan. 07, 2017 | Frank Obermüller

Results Comparison: System Value vs. Market Value Regional System Value of Wind (Nodal pricing) Regional Market Value of Wind (Uniform pricing)     Regional market value of wind (based on zonal pricing, no grid situation included) Value_loc = p*g_loc / p*1 Finding: Differences occur Reason: Different production profiles which are correlated differently as to the market price profile The regional value of wind is strongly dependent on the pricing regime (and thus on the grid situation) Regional wind values | Jan. 07, 2017 | Frank Obermüller

Results Comparison: Regional market profits   Market profit under Nodal pricing Market profit under Uniform pricing Can be considered as efficient benchmark since grid situation is internalized to market profits Uniform pricing may favor regions which are not system optimal and, e.g., increase grid congestions Lot of the values are close to each other Some values have great differences In South: Regional value under nodal pricing is higher In North: lower. This is distortion between market value with and without grid side internalized. The pricing regime (nodal vs. uniform) influences the regional market profits. Under uniform pricing, inefficient allocations may be incentivized Regional wind values | Jan. 07, 2017 | Frank Obermüller

Results Comparison: Nodal market profit vs. Fixed FiT Market profit under Nodal pricing Wind profit of Fixed Feed-in Tariff (for 1 MW turbine, 1 year) (for 1 MW turbine, 20 years) Left: Wind profit: Higher in the North Lower in the South Right: Wind value Higher in the South Lower in the North Excactly opposite A Fixed Feed-in Tariff (without a regional component) does not sufficiently consider the regional value of wind and may favor different (sub-optimal) wind locations Regional wind values | Jan. 07, 2017 | Frank Obermüller

Conclusion Contribution to existing research: Quantification of (1) regional system values, (2) regional market values, and (3) fixed feed-in remunerations for German wind production Findings: The Regional System Value of Wind Production could have structural breaks, e.g., as in Germany The pricing regime (nodal vs. uniform) influences the regional market profits. Under uniform pricing, inefficient allocations may be incentivized A Fixed Feed-in Tariff (without a regional component) does not sufficiently consider the regional value of wind and may favor different (sub-optimal) wind locations Conclusion: A well-designed subsidy scheme needs to incorporate the regional system value This can be achieved, e.g., via nodal pricing and a market price component A fixed feed-in tariff may set wrong capacity allocation incentives Regional wind values | Jan. 07, 2017 | Frank Obermüller

Exploring Energy Markets – Enhancing Decisions! Frank Obermüller frank.obermueller@ewi.research-scenarios.de EWI – Institute of Energy Economics at the University of Cologne ewi Energy Research & Scenarios gGmbH Alte Wagenfabrik Vogelsangerstraße 321 D-50827 Cologne Regional wind values | Jan. 07, 2017 | Frank Obermüller

Backup / Appendix Regional wind values | Jan. 07, 2017 | Frank Obermüller

Results Current reimbursements of fixed feed-in tariff   Wind profit of Fixed Feed-in Tariff 1 Modelled wind production data per turbine 2 Calculated reimbursements based on Fixed Feed-in Tariff How calculated: modelled wind production data per turbine obtained from weather model Based on 20-year wind speed data Regional high resolution (6x6 km) Production per turbine averaged to one production year Calculated to eeg 2014 reimbursement (5+x) years starting reimbursement (X.X €cent/kWh) (15-x) base reimbursement (X.X €cent/kWh) X is dependent on the individual wind production compared to the wind production of a reference turbine (for 5 years) The lower the individual wind production (to reference), the longer the starting reimbursement Of an exemplary wind turbine: 1 MW Turbine 66m Hub-Height 2380 m2 swept area Regional wind values | Jan. 07, 2017 | Frank Obermüller