Title Page

Abstract

Résumé

Contents

List of Glossaries 17

Acronym & Abbreviation 18

Notation & Nomenclature 21

Chapter 1. Introduction: Electricity Demand and Role of Demand Response 25

1.1. Background 25

1.2. Theoretical Rationale for DR in Economics 28

1.3. Objectives and Scope of Research 31

1.4. Literature Review 34

1.4.1. Definition of DSM and DR 34

1.4.2. Relations between DSM, DR, Smart Grid, Renewable Energy, Climate Change, and Capacity Market 40

1.4.3. Advantages and Challenges of DR 41

1.5. Conclusion 47

Chapter 2. Case Studies: The U.S., France, and South Korea 48

2.1. Case of the U.S. 48

2.2. Case of France: NEBEF (La Notification d'Échange de Blocs d'Effacement) Mechanism 51

2.2.1. Background and Motivations to Introduce NEBEF in France 51

2.2.2. Introduction of NEBEF 58

2.2.3. Evolution of NEBEF 59

2.3. Case of South Korea: Demand Resource Trading Market (DRTM) 70

2.3.1. Motivations to Introduce DRTM 70

2.3.2. Recent Performance of DRTM in South Korea 74

2.4. Conclusions on Comparisons between France and South Korea 80

2.5. Conclusion 83

Chapter 3. Customer Baseline Load (CBL) Estimation 85

3.1. Importance of the Accurate CBL Estimation Methods and Difficulties 85

3.2. Methods to Calculate Customer Baseline Load 87

3.3. To Assess the Estimated CBL 89

3.4. CBL for an Average Household in South Korea 91

3.5. CBL for an Average Household in France 99

3.6. Conclusion 114

Chapter 4. Cost-Benefit Analysis of Demand Response 116

4.1. Necessity of Cost-Benefit Analysis 116

4.2. Costs & Benefits to Be Considered and Previous CBAs 117

4.3. CBA on South Korean Demand Resource Trading Market 120

4.3.1. A Model for the CBA 120

4.3.2. Decision-making Analysis for South Korean DRTM 135

4.4. CBA on the French NEBEF Mechanism 139

4.4.1. A Model for the Case of Load Shedding 141

4.4.2. Decision-making Analysis for the Case of Load Shedding Based on CBA 151

4.4.3. A Model for the Case of Load Shifting 153

4.4.4. Decision-making Analysis for the Case of Load Shifting based on CBA 156

4.5. Sensitivity Analysis 160

4.5.1. SA for the CBA of the South Korean Case 160

4.5.2. SA for the CBA of the French Case 172

4.6. Conclusion 195

Chapter 5. Conclusions and Implications 197

Bibliography 202

Appendices 211

Appendix A. Supplementary Figures & Tables 211

Appendix B. Résumé en français 252

Index 270

Table 1.1. Summary for the Sources of Datasets used in the Analyses of CBL... 32

Table 1.2. The Top 10 States in terms of the Number of Customers and AMI Pen-... 45

Table 2.1. The Monthly Reserved and Realized NEBEF Volume and Total in 2014 60

Table 2.2. The Monthly Reserved and Realized NEBEF Volume and Total in 2015 62

Table 2.3. The Monthly Reserved and Realized NEBEF Volume and Total in 2016 64

Table 2.4. The Size of the Fund for DR Program in South Korea from 2012 to 2016 74

Table 2.5. The Performance of South Korean DRTM in 2014, 2015, and 2016 76

Table 2.6. Comparison of the Electricity Market Structure between France and... 81

Table 2.7. The Comparison of the DR Mechanisms between France and South Korea 82

Table 3.1. The Top 5 Regions in terms of Number, Total Consumption, and Av-... 93

Table 3.2. Summary of the Established CBLs for the Event Times 97

Table 3.3. The CBL Accuracy Comparison (MAPE and RRMSE) among the CBLs... 98

Table 3.4. The Top 10 Département in terms of Number, Total Consumption,... 102

Table 3.5. Summary of the Established CBLs forrhe NEBEF (DR) Event Times 109

Table 3.6. The CBL Accuracy Comparison (MAPE and RRMSE) among the CBLs... 111

Table 3.7. Summary of the Established CBLs for the NEBEF (DR) Event Times... 112

Table 3.8. The CBL Accuracy Comparison (MAPE and RRMSE) between the Two... 112

Table 3.9. Overview of the Results: Accuracy of CBL Estimation Methods (MAPE... 114

Table 4.1. The Percentiles of the Cumulative Density Function of the Maximum... 121

Table 4.2. Descriptive Statistics of SMP in 2015, 2016 (South Korea) 128

Table 4.3. The SMP for the Event Times 129

Table 4.4. The Unit Retail Price for Residential Usage 130

Table 4.5. Different Time Periods for Peak and Off-Peak Periods (South Korea) 130

Table 4.6. ToU Tariff Scheme (South Korea) 130

Table 4.7. The Percentiles of the Cumulative Density Function of the Maximum... 140

Table 4.8. Descriptive Statistics of SMP in 2014, 2015, 2016 (France) 147

Table 4.9. The SMP for the DR Event (NEBEF) Times 148

Table 4.10. The Unit Average Retail Price for Residential Usage 148

Table 4.11. Summary for Results of Sensitivity Analyses 194

Table A.1. The Top 10 States in terms of the Number of Customers and AMI... 212

Table A.2. The Top 10 States in terms of the Number of Customers and AMI... 214

Table A.3. Descriptive Statistics of Loads in 2014, 2015, 2016 (France) 221

Table A.4. Descriptive Statistics of Loads in 2015, 2016 (South Korea) 232

Table A.5. The Performance on the Reductions of Electriciry Demand 235

Table A.6. The Composition of Participants by Industry 237

Table A.7. The Payments for the Capacity and Reductions of Electricity Demand 238

Table A.8. The Top 10 Département in terms of Number, Total Consumption,... 240

Table A.9. Summary of the Established CBLs for the Event Times (2016-01-21,... 243

Table A.10. The CBL Accuracy Comparison (MAPE and RRMSE) among the... 244

Table A.11. Summary of the Established CBLs for the NEBEF (DR) Event Times... 247

Table A.12. The CBL Accuracy Comparison (MAPE and RRMSE) among the... 248

Table A.13. Summary of the Established CBLs for the NEBEF (DR) Event Times... 248

Table A.14. The CBL Accuracy Comparison (MAPE and RRMSE) between the... 249

Table A.15. The Actual Load and Estimated CBL for the Event Day 250

Table A.16. The Required Load Reduction Level and the Target Reduction on... 250

Table A.17. The Target Load Reduction Level and the Target Reduction on 20%,... 251

Table B.1. Résumé des sources des données utilisées dans les analyses des... 254

Table B.2. Comparaisons des structures des marchés de l'électricité entres... 257

Table B.3. Comparaisons des mécanismes d'effacement entre France et... 258

Table B.4. Aperçu des résultats: précision des méthodes d'estimation des... 259

Table B.5. Résumé des résultats des analyses de sensibilité 263

Figure 1.1. Locations of Shin-Gori Nuclear Reactors No.5 & 6 Construction Site... 26

Figure 1.2. Timeline around Background and Introduction of South Korean DR Program 27

Figure 1.3. Inefficiencies of Average-Cost Pricing 28

Figure 1.4. Impact of DR in Regions with Organized Wholesale Markets 29

Figure 1.5. Impact of Demand Elasticity on Wholesale Price and Simplified Ef-... 29

Figure 1.6. Graphical Analysis of the DR in terms of Commercial (left) and Phys-... 30

Figure 1.7. Flow of the Research 33

Figure 1.8. Categorization of DSM 35

Figure 1.9. Categorization of DR Programs 37

Figure 1.10. Categorization of DR by NERC 38

Figure 1.11. Entities Play as Load Aggregators in Europe (previously and in 2014) 39

Figure 1.12. Smart electricity meter installations (2013-2017) 43

Figure 1.13. The Number of Residential Customers by State, 2016 in the U.S. 44

Figure 1.14. Residential Smart Meter Adoptation Rates by State, 2016 in the U.S. 44

Figure 2.1. DR Map of Europe as of 2013 52

Figure 2.2. DR Map of Europe as of 2014 53

Figure 2.3. DR Map of Europe as of 2017 54

Figure 2.4. France's Peak Demand over the Past Decade (Since 2007) 55

Figure 2.5. The Total Installed Capacity in 2015, 2016, and 2017 56

Figure 2.6. Heatmap of Loads, France in 2016 57

Figure 2.7. Timeline around Background and Introduction of French NEBEF 59

Figure 2.8. Load Duration Curve of France in 2014 (MW) Considering NEBEF... 61

Figure 2.9. Load Duration Curve of France in 2015 (MW) Considering NEBEF... 63

Figure 2.10. The Volume of Demand Reduction of France in 2016 (Programmes... 65

Figure 2.11. The Volume of Demand Reduction of France in 2016 (Chroniques... 66

Figure 2.12. Load Duration Curve of France in 2016 (MW) Considering NEBEF... 67

Figure 2.13. Comparing the Chronological Actualized NEBEF between 2014,... 68

Figure 2.14. The List of the NEBEF Operators 69

Figure 2.15. Heatmap of Loads, South Korea in 2016 70

Figure 2.16. Peak Load of South Korea from 2012 to 2016 71

Figure 2.17. Load Duration Curve of South Korea in 2016 72

Figure 2.18. Planned Capacity in 2017, 2022, and 2030 73

Figure 2.19. The Composition of Participants by Industry 77

Figure 2.20. The Payments for the Reduction and Capacity at South Korean... 78

Figure 3.1. One Demonstrative Customer Baseline Load (CBL) 85

Figure 3.2. The Number of Residential Customers by Region, 2016 in South Korea 92

Figure 3.3. The Total Consumption of Residential Customers by Region, 2016 in... 92

Figure 3.4. The Average Consumption of Residential Customers by Region, 2016... 93

Figure 3.5. The Visual Comparison among the CBLs Established with Different Methods 97

Figure 3.6. The CBL Accuracy Comparison (MAPE & RRMSE) among the CBLs... 98

Figure 3.7. The Number of Residential Customers by Département, 2016 in France 100

Figure 3.8. The Total Consumption of Residential Customers by Département,... 100

Figure 3.9. The Average Consumption of Residential Customers by Départe-... 101

Figure 3.10. The Visual Comparison among the Loads of the Previous Ten Days,... 106

Figure 3.11. The Visual Comparison among the Loads of the Previous Four... 108

Figure 3.12. Comparing the Actual Loads with all the Four CBLs 109

Figure 3.13. The CBL Accuracy Comparison (MAPE & RRMSE) among the CBLs... 110

Figure 3.14. Comparing the Actual Loads with the Two CBL Methods of CBLWMA....[이미지참조] 111

Figure 3.15. The CBL Accuracy Comparison (MAPE and RRMSE) between the Two... 112

Figure 4.1. Cumulative Density Function of the Maximum Reference Days (20... 121

Figure 4.2. The Remained Loads according to the Required Load Reduction... 122

Figure 4.3. SMP, South Korea in 2016 128

Figure 4.4. Raincloud plots of SMPs in 2015, 2016 (South Korea) 129

Figure 4.5. Actualized Load Comparison between without DR and with DR Par-... 137

Figure 4.6. Cost-Benefit plane: DR vs. without DR 138

Figure 4.7. Cumulative Density Function of the Maximum Reference Days (20... 140

Figure 4.8. The Remained Loads according to the Target Load Reduction Level... 141

Figure 4.9. SMP, France in 2016 146

Figure 4.10. Raincloud Plots of SMPs in 2014, 2015, 2016 (France) 147

Figure 4.11. Actualized Load Comparison between without DR and with DR... 152

Figure 4.12. Cost-Benefit Plane: DR vs. without DR (NEBEF, Load Shedding) 153

Figure 4.13. Actualized Load Comparison between without DR and with DR... 158

Figure 4.14. Cost-Benefit Plane: DR vs. without DR (NEBEF, Load Shifting) 159

Figure 4.15. 1,000 Time Simulation in terms of the Different Values of 'Cap'... 161

Figure 4.16. Sensitivity Analysis according to the Values of β (South Korean DRTM) 162

Figure 4.17. The Distribution of the θ Value for Each β Level (South Korean DRTM) 163

Figure 4.18. Sensitivity Analysis according to Change of Tariff Scheme (ToU)... 164

Figure 4.19. The Distribution of the θ Value for Each β Level according to... 165

Figure 4.20. Sensitivity Analysis accoridng to Change of Form of Function... 167

Figure 4.21. The Distribution of the θ Value for Each β Level accoridng to... 167

Figure 4.22. Highest & Lowest SMPs, South Korea in 2016 169

Figure 4.23. Sensitivity Analysis accoridng to Change of SMPs (Highest, South... 170

Figure 4.24. The Distribution of the θ Value for Each β Level accoridng to... 170

Figure 4.25. Sensitivity Analysis accoridng to Change of SMPs (Lowest, South... 171

Figure 4.26. The Distribution of the θ Value for Each β Level accoridng to... 171

Figure 4.27. Threshold SMP of Net Benefit for Each β Value (South Korean DTRM) 172

Figure 4.28. Sensitivity Analysis according to the Values of β for NEBEF (Load Shedding) 174

Figure 4.29. Sensitivity Analysis according to the Values of β for NEBEF (Load Shifting) 175

Figure 4.30. Sensitivity Analysis according to the Values of β for NEBEF (Load... 177

Figure 4.31. Sensitivity Analysis according to the Values of β for NEBEF (Load... 178

Figure 4.32. Sensitivity Analysis according to the Values of β for NEBEF (Load... 180

Figure 4.33. Sensitivity Analysis according to the Values of β for NEBEF (Load... 181

Figure 4.34. Sensitivity Analysis according to the Values of β for NEBEF (Load... 183

Figure 4.35. Sensitivity Analysis according to the Values of β for NEBEF (Load... 184

Figure 4.36. Sensitivity Analysis according to the Different Form of Function... 186

Figure 4.37. Highest & Lowest SMPs, France in 2016 187

Figure 4.38. Sensitivity Analysis according to the Change of SMPs (Highest,... 188

Figure 4.39. Sensitivity Analysis according to the Change of SMPs (Lowest, Load... 189

Figure 4.40. Threshold SMP of Net Benefit for Each β Value (Load Shedding,... 190

Figure 4.41. Sensitivity Analysis according to the Change of SMPs (Highest,... 191

Figure 4.42. Sensitivity Analysis according to the Change of SMPs (Lowest,... 192

Figure 4.43. Threshold SMP of Net Benefit for Each β Value (Load Shifting,... 193

Figure A.1. The Number of Commercial Customers by State, 2016 in the U.S. 211

Figure A.2. Commercial Smart Meter Adoptation Rates by State, 2016 in the U.S. 212

Figure A.3. The Number of Industrial Customers by State, 2016 in the U.S. 213

Figure A.4. Industrial Smart Meter Adoptation Rates by State, 2016 in the U.S. 213

Figure A.5. Heatmap of Loads, France in 2014 214

Figure A.6. Heatmap of Loads, France in 2015 215

Figure A.7. The Volume of Demand Reduction of France in 2014 (Programmes... 216

Figure A.8. The Volume of Demand Reduction of France in 2014 (Chroniques... 217

Figure A.9. For the Comparison between the two volumes of NEBEF (MW) in 2014 218

Figure A.10. For the Comparison between the two volumes of NEBEF (MW) in... 219

Figure A.11. Load Duration Curve of France in 2014 220

Figure A.12. Raincloud plots of Loads in 2014, 2015, 2016 (France) 222

Figure A.13. The Volume of Demand Reduction of France in 2015 (Programmes... 223

Figure A.14. The Volume of Demand Reduction of France in 2015 (Chroniques... 224

Figure A.15. For the Comparison between the two volumes of NEBEF (MW) in 2015 225

Figure A.16. Load Duration Curve of France in 2015 226

Figure A.17. For the Comparison between the two volumes of NEBEF (MW) in 2016 228

Figure A.18. For the Comparison between the two volumes of NEBEF (MW) in... 229

Figure A.19. Load Duration Curve of France in 2016 230

Figure A.20. Heatmap of Loads, South Korea in 2015 231

Figure A.21. Raincloud Plots of Loads in 2015, 2016 (South Korea) 233

Figure A.22. Load Duration Curve of South Korea in 2015 234

Figure A.23. The Reductions of the Electricity Demand at the South Korean... 236

Figure A.24. The Registered Capacity and the Number of Parricipanrs of the... 237

Figure A.25. The Number of Industrial Customers by Département, 2016 in France 238

Figure A.26. The Total Consumption of Industrial Customers by Département,... 239

Figure A.27. The Average Consumption of Industrial Customers by Départe-... 239

Figure A.28. SMP, South Korea in 2015 241

Figure A.29. SMP, France in 2014 241

Figure A.30. SMP, France in 2015 242

Figure A.31. The Visual Comparison among the CBLs Established with Different... 243

Figure A.32. The CBL Accuracy Comparison (MAPE & RRMSE) among the CBLs... 244

Figure A.33. The Visual Comparison among the Loads of the Previous Ten Days,... 245

Figure A.34. The Visual Comparison among the Loads of the Previous Four... 246

Figure A.35. Comparing the Actual Loads with all the Four CBLs (2016-07-19,... 246

Figure A.36. The CBL Accuracy Comparison (MAPE & RRMSE) among the CBLs... 247

Figure A.37. Comparing the Actual Loads with the Two CBL Methods of CBLWMA....[이미지참조] 248

Figure A.38. The CBL Accuracy Comparison (MAPE & RRMSE) between the Two... 249

Graphique B.1. Analyse graphique de l'effacement en termes d'électricité com-... 253

Graphique B.2. Déroulement de la recherche 255

Graphique B.3. Comparaison de charge réalisée entre sans et avec participation... 261

Graphique B.4. Analyse de sensibilité selon les valeurs de β (DRTM en Corée du Sud) 262

Box 4.1. Assumptions of 'Reference Scenario' (South Korean DRTM) 136

Box 4.2. Assumptions of 'Reference Scenario' (Load Shedding, French NEBEF) 151

Box 4.3. Assumptions of 'Reference Scenario' (Load Shifting, French NEBEF) 157

Box 4.4. Assumption Changes from 'Reference Scenario': Inconvenience Costs (South Korean DTRM) 162

Box 4.5. Assumption Changes from 'Reference Scenario': Tariff Scheme (South Korean DTRM) 164

Box 4.6. Assumption Changes from 'Reference Scenario': Different Form of Function coefINC(d) based on ToU...[이미지참조] 166

Box 4.7. Assumption Changes from 'Reference Scenario': Highest & Lowest SMPs (South Korean DTRM) 168

Box 4.8. Assumption Changes from 'Reference Scenario': Inconvenience Costs (French NEBEF) 173

Box 4.9. Assumption Changes from 'Reference Scenario': CBL (French NEBEF) 176

Box 4.10. Assumption Changes from 'Reference Scenario': Tariff Scheme (French NEBEF) 179

Box 4.11. Assumption Changes from 'Reference Scenario': CBL & Tariff Scheme (French NEBEF) 182

Box 4.12. Assumption Changes from 'Reference Scenario': Different Form of Function CoefINC based on ToU Tariff...[이미지참조] 185

Box 4.13. Assumption Changes from 'Reference Scenario': Highest & Lowest SMPs (French NEBEF) 187

Worldwide concern on CO₂ emissions, climate change, and the energy transition made us pay more attention to Demand-side Management (DSM). In particular, with Demand Response (DR), we could expect several benefits, such as increased efficiency of the entire electricity market, enhanced security of electricity supply by reducing peak demand, and more efficient and desirable investment as well as the environmental advantage and the support for renewable energy sources. In Europe, France launched the NEBEF mechanism at the end of 2013, and South Korea inaugurated the market-based DR program at the end of 2014. Among a number of economic issues and assumptions that we need to take into consideration for DR, Customer Baseline Load (CBL) estimation is one of the most important and fundamental elements. In this research, based on the re-scaled load profile for an average household, several CBL estimation methods are established and examined thoroughly both for Korean and French DR mechanisms. This investigation on CBL estimation methods could contribute to searching for a better and accurate CBL estimation method that will increase the motivations for DR participants. With those estimated CBLs, the Cost-Benefit Analyses (CBAs) are conducted which, in turn, are utilized in the Decision-making Analysis for DR participants. For the CBAs, a simple mathematical model using linear algebra is set up and modified in order to well represent for each DR mechanism's parameters. With this model, it is expected to provide an intuitive and clear understanding of DR mechanisms. This generic DR model can be used for different countries and sectors (e.g. residential, commercial, and industrial) with a few model modifications. The Monte Carlo simulation is used to reflect the stochastic nature of the reality and the optimization is also used to represent and understand the rationality of the DR participants, and to provide microeconomic explanations on DR participants' behaviors. In order to draw some meaningful implications for a better DR market design, several Sensitivity Analyses (SAs) are conducted on the key elements of the model for DR mechanisms.

이산화탄소배출, 기후변화, 에너지전환에 대한 전 세계적인 관심에 따라 수요관리(Demand-side Management, DSM)에 대한 관심이 더욱 증가하고 있다. 특히 수요반응(Demand Response, DR)은 전력시장 효율성 제고, 최대부하 감축을 통한 전력공급안보 강화, 더 효율적이고 바람직한 투자, 친환경 이점과 재생가능에너지에 대한 보완 등과 같은 많은 편익을 가져올 것으로 예상한다. 유럽에서 프랑스는 수요반응 프로그램 NEBEF(La Notification d'Échange de Blocs d'Effacement)를 2013년 말부터 운영하기 시작하였고, 한국에서는 2014년 말부터 시장기반 수요반응 프로그램인 수요자원거래시장을 운영하기 시작하였다. 수요반응과 관련하여 고려해야 할 수 많은 경제학적 주제 및 가정 중에서, 고객기준부하(Customer Baseline Load, CBL) 추정은 가장 중요하고도 근본적인 요소 중 하나이다. 이 연구에서는 평균적인 가계를 위해 재산출한 수요 정보를 기초로, 한국과 프랑스 수요반응 기제에서 사용되는 고객기준부하 추정 방법을 비교·고찰한다. 궁극적으로 보상에 영향을 미치는 고객기준부하 추정 방법에 대한 심도 깊은 분석은 더 정확한 고객기준부하 추정 방법을 탐색함으로써 수요자원거래시장 참여 동기를 제고할 것으로 기대한다. 검토한 고객기준부하 추정 방법을 바탕으로, 비용편익분석(Cost-Benefit Analysis, CBA)을 실시하고 이를 다시 수요자원거래시장 참여에 대한 의사결정분석(Decision-making Analysis)에 사용한다. 비용편익분석을 위해 선형대수(Linear Algebra)를 이용한 간단한 수학적 모형은 프랑스·한국 각각의 수요반응 기제의 파라미터들을 잘 반영하도록 고안하였다. 이 모형은 수요반응 기제에 대한 직관적이고 명확한 이해를 제공할 것으로 기대한다. 고안한 모형을 조금만 수정하면 주택 부문 이외에 상업·산업 부문 또는 다른 나라의 수요반응에도 적용이 가능하다. 현실의 임의적이고 우발적인 속성을 잘 반영하기 위해 몬테카를로 시뮬레이션(Monte Carlo Simulation)을 사용하였고, 수요자원거래 시장 참여자들의 경제학적 합리성을 반영·이해, 그리고 미시경제학적 설명을 제공하기 위해 최적화(Optimization)를 사용하였다. 더 나은 수요자원거래시장 설계를 위한 의미있는 정책 함의를 도출하기 위하여, 수요반응 기제의 핵심 요소들에 대한 민감도 분석(Sensitivity Analysis, SA)을 실시하였다.