Stochastic Optimization in Electricity Market Modelling The Effects - - PowerPoint PPT Presentation

Stochastic Optimization in Electricity Market Modelling

The Effects of Uncertainties in CO2 Certificate Prices and Network Expansion Costs

Lena Kittel Veronika Grimm Mirjam Ambrosius Gregor Zöttl Vienna September 6, 2017

Research Question

How do uncertainties in the regulatory framework afgect the investment decisions of transmission operators and private generators in the German Electricity Market? What are the welfare efgects of taking uncertainty into account?

Lena Kittel 2

Roadmap of this talk

• 1. Underlying electricity market model
• 2. Modelling of regulatory uncertainties
• 3. Computational results
• 4. Conclusion and outlook

Lena Kittel 3

• 1. Underlying electricity market model

Lena Kittel 4

Network expansion (social planner) Investment in generation facilities Trading at spot markets (competitive companies) Cost based redispatch (social planner)

• 1. Stage: Spot market and generation

investments

• 2. Stage: Redispatch and grid expansion

max firms’ profits (firms) s.t. Generation investment Production constraints Kirchhoff’s 1st law Flow restrictions max social welfare (TSO) s.t. Kirchhoff’s 1st and 2nd law Phase angels Power flow constraint Generation capacities Interconnection between spot market and redispatch amounts Variable constraints

Roadmap of this talk

• 1. Underlying electricity market model
• 2. Modelling of regulatory uncertainties
• 3. Computational results
• 4. Conclusion and outlook

Lena Kittel 5

Time schedule – Uncertain CO2 certificate prices

Lena Kittel 6

Numeric implementation – Uncertain CO2 certificate prices

Spot market: max E[firms’ profits] Redispatch: max E[social welfare] Influenced parameters and variables:

• Variable production costs

=> Spot market prices => Demands and generation amounts

• Variable production costs

=> Redispatch of generation quantities => Load shedding quantities => Prices => Generation and demand after redispatch Uninfluenced parameters and variables: => Generation investment costs and decisions => Renaturation decision => Line investment costs and decisions => Generation investment costs and decisions => Renaturation decision All Restrictions have to be fulfilled for all scenarios!

Lena Kittel 7

Time schedule – Uncertain future network expansion costs

Lena Kittel 8

Numeric implementation – Uncertain future network expansion costs

Spot market: max firms’ profits Redispatch: max E[social welfare] Influenced parameters and variables: => Spot market does not account for any transmission constraints

• Future network expansion costs

=> Future line investment decision => Redispatch of generation quantities ⇒ Load shedding quantities => Generation and demand after redispatch Uninfluenced parameters and variables: => Spot market prices => variable production costs and generation decisions => Investment costs and decision of private firms => Renaturation decision => Line investment decision of today All Restrictions have to be fulfilled for all scenarios!

Lena Kittel 9

Roadmap of this talk

• 1. Underlying electricity market model
• 2. Modelling of regulatory uncertainties
• 3. Computational results
• 4. Conclusion and outlook

Lena Kittel 10

Scenarios: CO2 certificate price uncertainties

Test scenario

Uncertain Case: Reference Case:

Deterministic CO2 certificate price:

40 €

Lena Kittel 11

Possible future CO2 certificate prices 30 € 40 € 50€ Probability weights 1/3 1/3 1/3

Results: CO2 certificate price uncertainties

Investment decisions of private firms: Ø No efgects in expanding generation capacities Investment decisions of the grid operator: Ø Additional line from Mecklenburg-West Pomerania to Saxony-Anhalt Ø Redispatch costs about 3.3% lower compared to status quo Social Welfare: + 3.15% compared to status quo Aggregated CO2 emissions: + 43,0% compared to status quo

Lena Kittel 12

Reference case ∆ lignite ∆ hard coal ∆ gas Uncertain case ∆ lignite

Uncertain future network expansion costs

Cost estimations from the german network development plan 2030: Reference Case: Expected value => Accounting for network uncertainties had negligibly small impacts on all results!

Lena Kittel 13

DC overhead line DC earth cable Cable per km 1.5 m. € 4.0 m. € DC converter station 0.2 m. € 0.2 m. € Lifetime 40 years 40 years Probability Weights 50 % 50 %

Roadmap of this talk

• 1. Underlying electricity market model
• 2. Modelling of regulatory uncertainties
• 3. Computational results
• 4. Conclusion and outlook

Lena Kittel 14

• 4. Conclusion and outlook

We have established an electricity market model which takes uncertainties into account. This framework allows to explicitly analyze the impact of different forms

• f regulatory uncertainties on general welfare and on generation and

network investment decisions. First qualitative results based on test example show that taking uncertain CO2 certificate prices into account has significant effects. Future work: analyze further scenarios, more decision stages, allow adjustments of the decisions

Lena Kittel 15

Thank you for your Attention!