Industrial Demand Response: Methods, best practices, case studies, and applications

Cover
Contents
About the editors
Foreword
Introduction
1 A comprehensive review on industrial demand response strategies and applications
	1.1 Introduction
	1.2 Demand side management and ancillary services in smart grid
		1.2.1 Smart grid
		1.2.2 Demand response automation schemes
		1.2.3 Ancillary services in the industrial sector
	1.3 Industrial DR case study implementations
		1.3.1 Manufacturing processes
		1.3.2 Refrigerator warehouses
		1.3.3 IT industry/data centers
	1.4 Barriers and limitations
		1.4.1 Financial
		1.4.2 Behavioral/social
		1.4.3 Regulatory
		1.4.4 Technological
	1.5 Conclusions
	References
2 Demand response cybersecurity for power systems with high renewable power share
	2.1 Introduction
	2.2 An overview of DR and EV-based DR
	2.3 An overview of demand side cybersecurity
	2.4 Modeling power system with DR
	2.5 Discussions on the results of cyberattacks on EV aggregator
	2.6 Conclusion
	References
3 Recurrent neural networks for electrical load forecasting to use in demand response
	3.1 Introduction
	3.2 DR programs
		3.2.1 Load forecasting in DR
	3.3 Review on load forecasting
	3.4 RNNs in electric load forecasting
		3.4.1 Scaling data, normalizing
	3.5 PCA for electrical load forecasting
	3.6 Load data pre-processing with time organization and training, validation and testing: case study of Urban Area of New South Wales
	3.7 Results and discussion
	3.8 Conclusion
	References
4 Optimal demand response strategy of an industrial customer
	4.1 Demand side management categories
	4.2 What is DR?
	4.3 Why DR?
	4.4 DR classification
		4.4.1 Competitive DR
		4.4.2 Non-competitive DR
		4.4.3 Incentivebased DR
	4.5 Benefits of DR
	4.6 Challenges in DR implementation
	4.7 DR provisions
	4.8 Applications of DR
	4.9 Motivation about DR
	4.10 DR of an industrial buyer
	4.11 Problem formulation
		4.11.1 Market clearing sub-problem
		4.11.2 Proposed purchase cost-saving optimization sub-problem
	4.12 Proposed solution algorithm
	4.13 Case study
	4.14 Conclusion
	References
5 Price-based demand response for thermostatically controlled loads
	5.1 Demand response
	5.2 Smart grid control
	5.3 Modeling of thermostatically controlled loads (TCL)
	5.4 DR from aggregated TCLs—load model
		5.4.1 Transfer function of aggregated response of TCL units
	5.5 Automatic generation control (AGC)
		5.5.1 Primary frequency control
		5.5.2 Secondary frequency control
	5.6 Dynamic demand control (DDC)
	5.7 Simulink model
	Appendix A: Modeling of aggregated TCL loads using coupled Fokker–Planck equations
	Appendix B: Single areaAGC system parameters
	References
6 Electric vehicle massive resources mining and demand response application
	6.1 Introduction
	6.2 Development status and trend of EVs and charging infrastructure
		6.2.1 Development status of EVs
		6.2.2 Construction situation of charging infrastructure
		6.2.3 Governments’ supporting policies
	6.3 EV massive resources digging and DR capability/potential evaluation
		6.3.1 EV massive resources digging
		6.3.2 EVs in DR capability/potential evaluation
	6.4 The mode of EVs participating in DR
		6.4.1 Research on multi-station mode participating in power grid DR
		6.4.2 Research on single-station mode participating in power grid DR
	6.5 Practical experience on EVs participating in DR
		6.5.1 DR pilot projects – in structural mode
		6.5.2 DR pilot projects – in event mode
		6.5.3 Practical experience
	6.6 Summary and prospect
		6.6.1 Summary
		6.6.2 Prospect
	References
7 Demand response measurement and verification approaches: analyses and guidelines
	7.1 Introduction
		7.1.1 Concepts
		7.1.2 Literature review
		7.1.3 Classification of CBL estimation methods
		7.1.4 Features of CBL estimation methods
	7.2 An overview of different CBL estimation approaches
		7.2.1 Averaging method
		7.2.2 Regression method
		7.2.3 Other CBL calculation methods
	7.3 Comparison of different baseline estimation methods
	7.4 Accuracy evaluation indexes
		7.4.1 RMSE and RRMSE
		7.4.2 MAPE and MAE [57, 58]
	7.5 Guidelines and suggestions to select a proper baseline estimation method
	7.6 Practical results
	7.7 Concluding remarks and outlook
	Acknowledgments
	References
8 Transactive energy industry demand response management market
	8.1 Demand response
	8.2 Transactive control
	8.3 DR modeling and simulation results
		8.3.1 Model A
		8.3.2 Model-B
		8.3.3 Simulation results and discussion
	8.4 TE management
	8.5 Methodology
		8.5.1 Bidding/offering strategy of energy storage devices (ESD)
		8.5.2 Bidding strategy of HVAC
		8.5.3 Offering strategy of PVs
	8.6 Problem formulation
	8.7 Simulation results and discussions
	8.8 Future works
	8.9 Conclusion
	References
9 Industrial demand response opportunities with residential appliances in smart grids
	Nomenclature
	9.1 Introduction
	9.2 Demand peaks
	9.3 Demand response
	9.4 Thermostatically controlled loads (TCLS)
	9.5 Case study 1: hybrid control approach for frequency regulation
		9.5.1 Refrigerator modelling
		9.5.2 DR controller description
		9.5.3 HillClimbing method
		9.5.4 System description
		9.5.5 Simulation results
		9.5.6 Discussion
	9.6 Case study 2: appliance level data analysis of summer demand reduction potential from residential aircons
		9.6.1 Summer peak demand analysis
		9.6.2 DR opportunities with aircons
	9.7 Conclusion
	References
10 Modelling and optimal scheduling of flexibility in energy-intensive industry
	10.1 Introduction
	10.2 Understanding flexibility across electricity consumer sectors
	10.3 Basis for an industrial flexibility model
		10.3.1 European grid balancing services
		10.3.2 Models in contemporary research
	10.4 Modelling framework formulation
		10.4.1 Definitions
		10.4.2 Modelling blocks
	10.5 Case study
		10.5.1 Model
		10.5.2 Results
	10.6 Conclusions
	Acknowledgements
	References
11 Industrial demand response: coordination with asset management
	11.1 Introduction
	11.2 Proposed strategy
		11.2.1 General idea
		11.2.2 Problem formulation
		11.2.3 Solution methodology
	11.3 Case study
		11.3.1 System description
		11.3.2 Results
		11.3.3 Discussion
	11.4 Conclusions
	11.5 Nomenclature
		11.5.1 Indices
		11.5.2 Parameters
		11.5.3 Variables
	11.6 Appendix
	References
12 A machine learning-based approach for industrial demand response
	12.1 Introduction
	12.2 Industrial load
		12.2.1 Characteristics of industrial load
	12.3 Industrial DR
		12.3.1 Industrial load forecasting
		12.3.2 Role of technology in IDR
		12.3.3 Role of policy in IDR
		12.3.4 Incentives and price-based DR
		12.3.5 Ancillary services
	12.4 Machine learning in IDR
		12.4.1 Genetic algorithm (GA)
		12.4.2 Support vector machine (SVM)
		12.4.3 Artificial neural network (ANN)
		12.4.4 Fuzzy logic
		12.4.5 Adaptive neuro-fuzzy inference system (ANFIS)
		12.4.6 Linear regressions
	12.5 Conclusion
	References
13 Feasibility assessment of industrial demand response
	13.1 Cost assessment of IDR
		13.1.1 Measurement of flexibility potential
		13.1.2 Design and deployment
		13.1.3 Operation and management
		13.1.4 Communication and control
		13.1.5 Feedback system
	13.2 IDR benefits
		13.2.1 Regulation services
		13.2.2 Reserves
		13.2.3 Self-consumption
		13.2.4 Changes in energy purchasing and flexibility trade
		13.2.5 Transmission and distribution network support
		13.2.6 Other benefits
	13.3 Feasibility assessment
		13.3.1 Indicators
	13.4 Case studies
		13.4.1 Chlor-alkali production industry
		13.4.2 Paper industry in Germany
	13.5 Conclusions and final considerations
	References
14 Measurement and verification of demand response: the customer load baseline
	14.1 Introduction
	14.2 Literature review
	14.3 Customer baseline load, non-intrusive load monitoring and physical-based load models
		14.3.1 The necessary linkage between DR methodologies
		14.3.2 Physical-based load models
		14.3.3 Unadjusted customer baseline load: a review of the main methodologies
		14.3.4 Adjustment coefficients for CBL
	14.4 Case study
		14.4.1 Detecting pre-heating and gaming through PBLM and NIALM
	14.5 Results and discussion
		14.5.1 Comparisons of unadjusted CBLs based on historical data
		14.5.2 Adjustment coefficients: weather sensitive (WS) and PBLM
		14.5.3 DR control events: effects on energy calculations
	14.6 Conclusions
	Acknowledgements
	References
15 Modeling and optimizing the value of flexible industrial processes in the UK electricity market
	15.1 Introduction
		15.1.1 Decarbonization challenges and value of demand response
		15.1.2 Industrial DR: significance and relevant work
		15.1.3 Chapter motivation and contributions
		15.1.4 Chapter outline
	15.2 Modeling framework
		15.2.1 Assumptions and generic formulation of industrial consumer’s optimization problem
		15.2.2 Uninterruptible processes with fixed power
		15.2.3 Interruptible processes with fixed power
		15.2.4 Uninterruptible and interruptible processes with discretely adjustable power
		15.2.5 Uninterruptible and interruptible processes with continuously adjustable power
		15.2.6 Material storage buffers
	15.3 Case study
		15.3.1 Description and input data
		15.3.2 Benefits of flexibility types with fixed power
		15.3.3 Benefits of flexibility types with adjustable power
		15.3.4 Benefits of material storage buffers
		15.3.5 Summary of benefits of different flexibility types
	15.4 Conclusions and future work
	Acknowledgement
	References
16 Case study ofAran Islands: optimal demand response control of heat pumps and appliances
	16.1 Origins of demand response programmes
		16.1.1 Traditional (industrial) DR applications
		16.1.2 Transition towards the residential sector
	16.2 RESPOND control loop and methodology
		16.2.1 IoT backend platform
		16.2.2 Forecasting services
		16.2.3 Optimisation services
		16.2.4 Control services
	16.3 Use case setup
		16.3.1 Pilot installations
		16.3.2 User interface
	16.4 Case studies and assessment
		16.4.1 Test case #1
		16.4.2 Test case #2
		16.4.3 Test case #3
		16.4.4 Test case #4
	Conclusion
	Acknowledgement
	References
17 Use case of artificial intelligence, and neural networks in energy consumption markets, and industrial demand response
	17.1 AI in energy market
	17.2 NN
	17.3 Power consumption and importance of its prediction
	17.4 Cogeneration and dual fuels
	17.5 DR and its importance
	17.6 Power consumption prediction using artificial NNs (ANNs)
	17.7 Framework of the NN-LSTM-based model
		17.7.1 Shell layer
		17.7.2 Input layer
		17.7.3 Hidden layer
		17.7.4 Attention layer
		17.7.5 Output layer
	17.8 Use case of NNs
		17.8.1 Overview and benefit
	17.9 RNN or LSTM:Which one is better for prediction?
		17.9.1 Overview
	17.10 Quantum technology
		17.10.1 Quantum computing
		17.10.2 Quantum fundamentals
	17.11 Quantum technology general applications
	17.12 Quantum technology and smart grids
	17.13 Forecasting in smart grids using quantum technology
	17.14 Final overview and conclusion
	Acronyms
	References
Index
Back Cover