Industrial Demand Response: Methods, best practices, case studies, and applications
2: University of Córdoba, Spain
3: Smart Grids and Smart Cities Laboratory, University of Salerno, Italy
Demand response (DR) describes controlled changes in the power consumption of an electric load to better match the power demand with the supply. This helps with increasing the share of intermittent renewables like solar and wind, thus ensuring use of the generated clean power and reducing the need for storage capacity.
This book conveys the principles, implementation and applications of demand response. Chapters cover an overview of industrial DR strategies, cybersecurity, DR of industrial customers, price-based demand response, EV, transactive energy, DR with residential appliances, use of machine learning and neural networks, measurement and verification, and case studies in the Aran Islands, as well as a use case of AI and NN in energy consumption markets.
The chapters have been written by an international team of highly qualified experts from academia as well as industry, ensuring a balanced and practically oriented insight. Readers will be able to develop and apply DR strategies to their respective systems.
Industrial Demand Response: Methods, best practices, case studies, and applications is a valuable resource for researchers involved with regional as well as industrial power systems, power system engineers, experts at grid operators and advanced students.
Inspec keywords: demand side management; smart power grids; energy resources; electric vehicles; learning (artificial intelligence); renewable energy sources; power markets; energy storage; power engineering computing; power supply quality
Other keywords: learning-artificial intelligence; demand side management; energy resources; power grids; renewable energy sources; power supply quality; power markets; electric vehicles; smart power grids; energy storage
Subjects: Automobile electronics and electrics; Power engineering computing; Other energy storage; Power system management, operation and economics; Power supply quality and harmonics
- Book DOI: 10.1049/PBPO215E
- Chapter DOI: 10.1049/PBPO215E
- ISBN: 9781839535611
- e-ISBN: 9781839535628
- Page count: 440
- Format: PDF
-
Front Matter
- + Show details - Hide details
-
p.
(1)
-
1 A comprehensive review on industrial demand response strategies and applications
- + Show details - Hide details
-
p.
1
–23
(23)
One of the key points marking the transition from traditional toward smart energy grids is the provision of flexibility services from the demand side. Power flexibility facilitates the integration of Renewable Energy Resources (RESs), while the balance between the supply and the demand side is maintained. Demand Response (DR) techniques are providing the opportunity for high exploitation of the flexibility potential, since they enable reducing, increasing or shifting a portion of the electrical demand, for a specific time period. The industrial sector is expected to have a more significant contribution compared to the residential and the commercial sector, mainly because of the high-consuming equipment, the scheduled operations and the already installed metering equipment on the facilities. The present work explores the state-of-the-art DR applications implemented in the industrial sector. The individual characteristics of each type of industry are analyzed, as they play a major role in the identification of DR potential, since industrial processes may involve critical loads, being highly correlated, that must follow strict operational constraints. The current level of participation in industrial DR programs is being assessed, identifying possible technical or regulatory limitations that prevent further adoption.
-
2 Demand response cybersecurity for power systems with high renewable power share
- + Show details - Hide details
-
p.
25
–39
(15)
Cybersecurity is crucial for modern power systems due to their high digitization. The open information and communication technology, which is (ICT) being used for the operation of such systems, is highly vulnerable to cyber threats. The adopted smart grid concept around the globe enables the utilization of demand-side for providing ancillary services based on well-known demand response (DR) programs. These programs aggregate smart appliances in homes and electric vehicles (EVs) for providing vital services such as frequency regulation and voltage support. Since the aggregation is based on the cyber layer, any cyber threat could affect the ancillary services that are being delivered from the aggregators, which might lead to stability and security issues resulting in brownout or massive blackouts. This chapter discusses the cybersecurity in DR program and shows its importance for modern and future smart power systems due to their stability and security margins. Furthermore, the cyberattack case study is implemented in a power system with a demand side program responsible for providing primary frequency support ancillary service, where the results confirm the high vulnerability of modern power systems to cyber threats on DR-active power reserve providers. Moreover, technical suggestions are provided for enhancing the cybersecurity in DR programs in power systems with high power share from renewable energy sources.
-
3 Recurrent neural networks for electrical load forecasting to use in demand response
- + Show details - Hide details
-
p.
41
–58
(18)
Electric load forecasting is a fundamental technique to understand end-user behavior and therefore a crucial factor in the design of demand response (DR) programs. Load forecasting will also identify the appropriate design of DR programs. In this chapter, a range of different machine learning applications are studied to represent the influential factors for electrical load demand forecast in a DR context, with a variety of different data scenarios, temporal and technical scenario. This chapter explores and compares the load prediction analysis through basic recurrent neural networks (RNNs); Vanilla RNN, gated recurrent units (GRU), and long short-term memory (LSTM), using principal component analysis (PCA). It is found that PCA can be used to reduce the number of principal components for Vanilla RNN, GRU, and LSTM networks. Reducing the number of principal components using PCA is one of the techniques that is used in dimensionality reduction. Reduction in dimensionality will relieve the computational burden. In this work, the dimensionality reduction improves the predictive output. It is observed that for electric load demand forecasting, the preferred technique is GRU, trained with a principal component. The performance is evaluated through mean absolute percentage error (MAPE), which is relatively lower than other techniques.
-
4 Optimal demand response strategy of an industrial customer
- + Show details - Hide details
-
p.
59
–75
(17)
Demand response (DR), which is an important feature of the smart grid, can play a vital role by making the demand side more responsive to the varying gap between demand and supply. DR is utilized by power utilities to maintain system reliability, security and stability while customers utilize it to reduce the electricity cost by increasing or decreasing the load during valley or peak demand periods. Industries consume huge amounts of electricity; therefore, DR strategies are required to be implemented by industrial customers to enhance the saving. Further, industrial customers can provide DR by employing many different technologies or strategies to achieve shifts in demand in the following ways: (i) reducing or interrupting consumption temporarily with no change in consumption in other periods, (ii) shifting consumption to other time periods, and (iii) temporarily utilizing onsite generation in place of energy from the grid.
-
5 Price-based demand response for thermostatically controlled loads
- + Show details - Hide details
-
p.
77
–102
(26)
Smart grid enables active participation of consumers daily operation of the grid through Demand Response (DR). DR refers to the actions initiated from contracted customers by changing their demand in response to price signals, incentives, or directions from grid operators. In this chapter, industrial DR suitable for frequency regulation is discussed. For this, a mathematical model of price-based DR from thermostatically controlled loads (TCL) for controlling the temperature of the chillers in large academic complex environment is presented. A probabilistic model of the density function of aggregated TCL loads is discussed. The variation of the thermostat set point demand temperature an increase in the price is presented. In order to match the power demand and power supply, a new method for dynamic demand control (DDC) with automatic generation control (AGC) in smart grid environment is proposed. A load frequency control using DDC was modeled in this study. The load frequency control model was simulated for a step load change of 0.01. The frequency deviation was compared with the frequency deviation obtained when generation control, using PI controller, alone was implemented for frequency control. Thus, DDC alone is required to maintain the system frequency, during small load variations. DDC will play a major role in reducing these losses caused to the GENCOs under a Smart Grid environment.
-
6 Electric vehicle massive resources mining and demand response application
- + Show details - Hide details
-
p.
103
–131
(29)
In 2020, the sales volume of electric vehicles (EVs) in China reached 1.367 million. A rapid growth trend was witnessed by the huge increment of electric vehicles in past the several years. By the end of 2020, China has nearly 5 million new energy vehicles. Meanwhile, China's charging infrastructure has reached 1,681,000 units. It is expected that the global penetration rate of new energy vehicles will exceed 30% in 2030. At that time, the number of electric vehicles in China is prospected to be 80-100 million. As the largest EV market in the world, China has unique conditions to develop and study the interactive application of EVs and power grid. The power system can have a chance of promoting comprehensive innovation thanks to the booming of the EV industry. A smart energy transportation network, that can participate in the grid demand response (DR) timely, would possibly consist of massive EVs, the power grid, renewable energy network and transportation network.
Because of the inherent mobile energy storage characteristics of EVs, flexible large-scale EV clusters have great potential in power load regulation, renewable energy consumption, power quality improvement, etc. Thus EVs can be used to participate in auxiliary services such as peak shifting and valley filling, frequency regulation, emergency support so as to interact friendly with the grid. In recent years, many cities in China have tried to include EVs in the pilot and made positive exploration in vehicle network interaction.
-
7 Demand response measurement and verification approaches: analyses and guidelines
- + Show details - Hide details
-
p.
133
–161
(29)
Demand response (DR) programs are defined as the ability of customers to change their consumption pattern in response to market/system signals. Nowadays, DR programs are interested worldwide as an essential part of the future power system and also considered as virtual generation resources. However, an accurate measurement and verification (M&V) approach is needed to implement these programs successfully. Indeed, the evaluation of the real potential of a DR program that is enabled during a DR event is depended on an evaluation method that should be employed to estimate the consumption behavior of the customers if they have not participated in DR. In this regard, customer base-load (CBL) estimation is defined as the approach to estimate the customers' load levels if they have not received DR calls. Then, by computing the difference among the estimated baseline and measured load data, the real potential of DR would be calculated. So, the determination of the real potential of DR is dependent on the difference between the estimated baseline and measured load data. Since various factors (such as load type, weather condition, day of a week, etc.) could affect the CBL, it is a challenging and complex task to provide an accurate estimation of the CBLs.
-
8 Transactive energy industry demand response management market
- + Show details - Hide details
-
p.
163
–186
(24)
In a smart grid paradigm, the concepts of demand response (DR) and transactive energy (TE) are used to optimize the consumption and generation in the power networks. In this chapter, two models for DR are analyzed based on the well-known Cobb-Douglas utility function. Both models maximize their utility, subject to different constraints. A time-of-use price-based DR program is employed. Restructuring in the electricity sector, with an increase in renewable energy resources and distributed energy management technologies, offers the potential for significant improvement in the efficiency of power systems through the TE framework. In a TE framework, prosumers of all sizes can participate in the double auction electricity markets via automated home energy management systems. Heating, ventilation, and air conditioning (HVAC) and energy storage devices are the two important loads in residential buildings that account for a large proportion of building energy consumption. A two-way exchange of energy and information is possible with the current advent of communication systems and net metering. In this work, we consider the case of solar photovoltaics (PV), HVAC, and energy storage devices (electric vehicles and battery energy storage systems) of prosumers participating in the retail real-time double auction market. The problem is formulated as maximization of social welfare subject to power balance and network constraints. Simulation studies and results are presented for the modified IEEE 13 node distribution system.
-
9 Industrial demand response opportunities with residential appliances in smart grids
- + Show details - Hide details
-
p.
187
–207
(21)
Supply-demand balance is imperative for the reliability of the power system. Inability to maintain this balance results in frequency deviation and system failure. The recent integration of renewable energy sources such as wind and solar have reduced inertia and variable output which leaves the power system at risk of disturbance while also reducing the controllability of generators. However, the latter-day demand response is coming across as an economical and effective way of adding to the reliability and security of power systems by managing the electricity demand of customers at times of severe power imbalance. This chapter carries out a detailed literature review of centralized and decentralized demand control approaches. As well as presents a novel demand control approach for providing frequency regulation by using domestic refrigerators as control loads. This chapter also carries out a detailed study of large-scale appliance level interval meter consumption data from Australia's largest network provider Ausgrid. Appliance level data is used in combination with household-level data to study the contribution of air-conditioners in summer peak demand. Clustering is performed on air-conditioner data to identify various air-conditioner load profile patterns. These patterns are then used with demand control strategies to study the possible load reductions from residential air-conditioner control across the Australian State of New South Wales.
-
10 Modelling and optimal scheduling of flexibility in energy-intensive industry
- + Show details - Hide details
-
p.
209
–239
(31)
Current environmental trends such as the rapid penetration of renewable energy resources (RES) and decommissioning of controllable but polluting generators are putting stress on the reliable operation of electricity systems. This reduction of flexibility and the increase in volatility in the supply-side call for compensation from other sources in the grid. Although the development of energy storage systems (ESS) is creating an opportunity to relax the energy balance constraints in the grid, it is currently not sufficient to solve the constantly growing need for flexibility.
-
11 Industrial demand response: coordination with asset management
- + Show details - Hide details
-
p.
241
–259
(19)
Demand response (DR) is one of the pillars of the modern distribution system, where consumers would voluntarily reduce consumption in response to financial incentives. While for residential consumers, demand curtailment is mainly a matter of inconvenience, for industrial customers, reduction in electric demand can lead to severe operational ramifications such as a halt in production, a pile-up of inventory, or wasted labor. These challenges have caused industrial DR to remain less explored compared to residential demand-side management. Although the industrial sector may be small by numbers, its energy consumption is the dominant load on most distribution systems. This further underlines the potential benefits gained by involving industrial loads in DR events.
-
12 A machine learning-based approach for industrial demand response
- + Show details - Hide details
-
p.
261
–274
(14)
Considerable interest is now being vested in low-carbon energy sources in other to meet the world's ever-growing energy demand without causing damage to the environment, which has given rise to the increasing contributions of renewable energy sources to the energy grid. This development is not without its challenges to modern electric power systems. Due to the intermittent nature of renewable energy resources, its increase has resulted in energy demand-supply mismatch, grid imbalance, or grid instability. A reliable and cost-effective approach is required to address this energy trade imbalance caused by the influx of renewable energy sources. Demand response (DR) is a concept that aims at achieving energy balance in the grid by controlling and adjusting flexible loads.
-
13 Feasibility assessment of industrial demand response
- + Show details - Hide details
-
p.
275
–300
(26)
As seen in this chapter, a cost-benefits analysis (CBA) is needed to measure IDR profitability. Such CBA should incorporate a LCC approach, which describe all cost involved in each stage of the IDR development cycle. Together with the cost, a specific analysis of the benefits should be conducted. As mentioned, such benefits are diverse and depend on technical requirements, standpoint (industrial customer, utility or aggregator), timelines (short, medium or long terms), market (wholesale, retail, balancing and flexibility) and pricing features. Finally, a proper choice of profitability indicators must be done, which allows to compare different IDR actions and select the most suitable and profitable option.
-
14 Measurement and verification of demand response: the customer load baseline
- + Show details - Hide details
-
p.
301
–327
(27)
Demand response (DR) is a basic tool to achieve power systems flexibility in the short and medium terms. The effective deployment of DR and the engagement of new resources need knowledge about how DR performs and how to evaluate their flexibility to give a correct economic feedback to customers and aggregators. DR verification requires a reference in the absence of control: the customer baseline load (CBL). The aim of this chapter is to describe several baselines that provide an acceptable evaluation of load response as well as the use of different adjustment methods to improve the CBL. Some of these adjustment factors can be justified through the simulation of physical-based load models (PBLMs), which are also used in DR for planning and operational tasks. The chapter discusses some issues reported by grid operators: detection of abnormal responses (before and after DR) that can be due to gaming or are reactions to maintain load service such as pre-heating, pre-cooling or the change of tasks timeline. All these approaches have been illustrated using real data of an industrial customer. Results show that the adjustment of CBLs can improve several conventional approaches described in the literature.
-
15 Modeling and optimizing the value of flexible industrial processes in the UK electricity market
- + Show details - Hide details
-
p.
329
–355
(27)
Despite its comparative advantages with respect to residential and commercial demand response (DR), industrial DR (IDR) in general, and modeling of different types of flexible industrial processes in particular, has received relatively limited research attention, with previous work having only explored limited and industry-sector-specific subsets of such processes. This chapter adopts an alternative, sector-agnostic modeling approach and develops generic models of all conceivable types of flexible industrial processes, with the aim to shed light on their key operating differences and assist industrial consumers interested in IDR schemes to identify and assess the types that are more relevant to their systems. In this context, this chapter identifies and discusses seven different types: (1) uninterruptible processes with fixed power, (2) interruptible processes with fixed power, (3) uninterruptible processes with discretely adjustable power, (4) interruptible processes with discretely adjustable power, (5) uninterruptible processes with continuously adjustable power, (6) interruptible processes with continuously adjustable power, and (7) material storage buffers.
-
16 Case study of Aran Islands: optimal demand response control of heat pumps and appliances
- + Show details - Hide details
-
p.
357
–377
(21)
Demand response has proven to be a crucial mechanism in the process of flexibility exploitation on the demand side. Throughout the years, demand response has evolved exploiting more and more previously untapped potential energy sources. In that process, residential users have provided a significant buffering capacity for balancing energy production and demand, but this came with a few challenges. With more and more households transitioning from being purely energy users to smart homes and energy prosumers with distributed renewable energy generation, new possibilities have opened up for integrated optimisation approaches that make the best use of both locally generated and grid-supplied energy as well as energy storage systems.
-
17 Use case of artificial intelligence, and neural networks in energy consumption markets, and industrial demand response
- + Show details - Hide details
-
p.
379
–399
(21)
Despite all achievements, and advances in energy markets, microgrids, and smart grids within the world, issues such as power distribution, consumption, or optimization are among the important and significant areas within the industry and technology. As industrialization and technology improve, these subjects become more important. Most of the experts attempt to have far better control on power consumption/distribution, and technologies like combined heat, and power (CHP), or gas-electricity, or demand forecasting, especially in smart sustainable cities (SSCs). Using artificial intelligence (AI) and neural networks (NNs) can have an important role in performing, and optimization that will lead to lowering the issues in future power systems. An NN-long short-term memory (LSTM)-based model can help the experts to control, predict, and optimize the facility consumption, and power distribution. Conceptually, in industrial and SSCs, more they develop, more the quantity of data is going to be generated that a simple and practical tool to research about and analyze these big data is AI. Regarding an outsized amount of data, the training and predicting process of AI is going to be far more accurate, due to the low root mean square error (RMSE). Accordingly, the result is going to be near the actual and help the SSCs to possess controlled power consumption, distribution, and CHPs. Also, the combination of quantum technology with smart grids, and NNs are analyzed. Accordingly, the mentioned technologies cause preventing power loss and promoting a way to a smarter, technology-based, and sustainable world with high ability of demand response (DR).
-
Back Matter
- + Show details - Hide details
-
p.
(1)