The electric power supply system serves a critical functionality in modern society. Its primary operational objective is to be able to match the generated power to the load demand at all times. Secure, reliable, economic, efficient and sustainable operation are other key objectives for power system operation. Traditionally, efforts to realize these technical and economic operational objectives have almost entirely concentrated on the supply side of the power system, that is, the whole electrical infrastructure that is involved in the generation, transmission and delivery of electrical power to the end-users [1]. Over the past few decades, however, attention has increasingly been given to the potential role that the demand side (i.e., the consumer of electricity) can play in the secure, reliable and efficient operation of the power system. The concept of demand-side management (DSM) largely originated in the U.S.A. in the 1970s, in response to growing energy security concerns as well as the environmental impact of electricity generation, especially nuclear power [2]. Its evolution was influenced by a combination of political, economic, social, technological and resource supply factors that significantly impacted the electricity sector’s operating environment and its outlook for the future [3]. The ensuing decades witnessed many developments, including the deregulation of the electric power supply industry that introduced competitiveness and a heightened focus on greater efficiency and economy of operation, the proliferation of decentralized and variable renewable energy generation integrated into the distribution network and technological advancements that spurred efforts to evolve the legacy electric grid into a modern smart grid ^[4][5][6][4,5,6]. In this context, DSM has been viewed as having the potential to play an increasingly important role in supporting the electric grid modernization efforts [7].

DSM refers to a variety of schemes and mechanisms, mainly implemented by the electric utility, meant to influence customers’ energy use patterns in an effort to achieve a more desirable utility’s load shape, in terms of the load’s time pattern and/or its magnitude [3]. Electric power systems in many parts of the world are faced with a variety of challenges. Growing energy demand coupled with the increasing cost of building new generation, transmission and distribution infrastructure is compelling many system operators to operate the power system closer to the system capacity limit. This leads to the system being overstressed and increases the cost of the electricity supply due to the need to frequently run expensive peaking generation [1]. The increased uptake of variable renewable generation at large scales due to environmental and energy security concerns is having a significant impact on the dynamics of power system operation, leading to a greater need for operational flexibility and control. DSM is anticipated to play a greater role as a flexible resource that can support the system operator in addressing these and other challenges to ensure the system is operated as reliably, efficiently and economically as possible [8]. DSM programs are mainly aimed at modifying the behavior of electrical loads of the various customer types (industrial, commercial and residential) in an effort to optimize energy production costs and enhance energy utilization, leading to improved system reliability. They may even enable the deferring of network reinforcement by diminishing the peak capacity requirements of the network when managed effectively [9].

2. Overview of Demand-Side Management (DSM)

2.1. Historical Development of DSM

Demand-side management, as an alternative to supply-side management for the purpose of achieving least-cost electrical energy provision, grew steadily in the U.S.A. in the 1970s out of increasing concern over the reliance on foreign sources of fossil fuels and awareness regarding the environmental consequences of electricity generation [2]. The growth and popularity of DSM programs among U.S.A. electric utilities were largely fueled by favorable regulatory policies that provided the regulatory environment and incentives that encouraged the pursuit of least-cost or integrated resource planning principles on the part of the utilities. The development and evolution of DSM over the decades has been highly influenced by political, economic and regulatory factors, as well as public–private interests of the stakeholders. Although the importance of influencing consumer energy use behavior (by means of the proper choice of energy-consuming devices and appliances and their appropriate usage) was long recognized, a holistic approach in the form of a DSM framework had not really been developed until the late 1970s. The development of this framework enabled addressing such questions as [3]:

• Considering the current set of resources available or under consideration, what changes in consumer demand patterns would be of benefit to the consumers and to suppliers?
• Which end-use technologies or changes in consumer behavior are likely to yield those changes?
• What market implementation methods would be needed to influence consumer preferences and behavior to produce the desired result?

Thus, rather than look at energy efficiency, load management and other related initiatives as independent tools, the DSM framework would provide a holistic and logical approach to implementing all initiatives meant to positively influence electricity consumption behavior in order to achieve the desired objectives. Early attempts to manage demand for electricity included the use of price as an incentive (i.e., time-of-use pricing), technologies such as thermal energy storage (i.e., storage water heating) and direct load control, being one of the most prevalent [10]. South Africa has historically enjoyed an abundant electric energy supply, but early in 2008, the country was faced with an acute power supply deficiency problem caused by a combination of supply-side problems that included coal availability, maintenance needs and unplanned outages, leading to a drastic fall in reserve margins. Energy efficiency and demand-side management measures were identified as an immediate and effective means of addressing the power supply inadequacy problem [11]. Indeed, a number of policies have been formulated in the recent past that were directly intended to address the issue of energy efficiency and demand-side management. For example, the “Regulatory Policy on Energy Efficiency and Demand Side Management for South African Electricity Industry” of 2004 [12] provided the regulatory framework for the implementation of energy efficiency and demand-side management as a means of addressing the anticipated peak generation capacity constraints. As the power supply shortages have intensified in recent times, Eskom (the national electric grid operator) has been placing greater emphasis on demand-side management, seen to be a “win-win situation—reducing pressure on the power system and enabling customers to realize cost savings by being more energy-conscious and reducing their consumption without affecting business productivity or quality of life” [13].

2.2. Implementation of DSM

DSM encompasses a variety of (largely) utility-centric initiatives that range from educating and sensitizing customers regarding the efficient and conscientious use of energy to offering a variety of incentives that are aimed at realizing an improved load shape. The main categories of DSM programs are [2]:

• General information educating customers regarding the effective use of energy
• Advisory technical support, including energy auditing and making recommendations for improvement in energy use
• Price-based DSM schemes attempt to induce changes in customer energy usage patterns by means of changes in the price of electricity. This encompasses real-time pricing, critical peak pricing and time-of-use rates. The premise for the use of price-based DSM schemes is that if the price differentiation for different time periods is significant enough, the customers are likely to respond by altering their energy use so as to take advantage of lower-priced periods, thereby reducing their energy bills. Of course, customers may also choose to continue with their normal routine of energy consumption, so participation in this type of DSM is categorized as voluntary.

2.2.1. Incentive-Based Schemes

2.4. Benefits of DSM

The electricity supply sector has many stakeholders, including transmission and distribution system operators, electricity market operators and balance responsible parties (BRP), energy service providers, energy traders, suppliers, traders and retailers, manufacturers, customers, policymakers and regulators, among others. All these stakeholders can derive benefits from the effective implementation of demand-side management:

• Transmission and distribution system operators can take advantage of the flexibility provided by the demand response to realize improved system stability in the face of the higher penetration of variable renewable generation, improved congestion management and a decrease in network bottlenecks that may be caused by sustained peak demand. Demand-side management may also contribute to improved voltage regulation and overall power quality
• Customers may derive economic benefits from demand-side management, which makes a variety of choices available for better managing energy usage, as well as actively participating in network management by means of taking part in various system operator-driven demand-side management initiatives
• Manufacturers may benefit from the opportunity provided by demand-side management to develop new products and technologies in support of the effective implementation of demand-side management
• Policymakers and regulators have a major role to play in the realization of demand-side management initiatives. They may also benefit when the successful implementation of these initiatives leads to technical, economic, social and environmental benefits, which is their ultimate goal
• Demand-side management may also offer an opportunity for new entrants in the electricity supply sector, as new and innovative paradigms are discovered for the most effective means of exploiting demand-side resources. For example, aggregators, advanced ICT and advanced metering infrastructure service providers will play an increasingly greater role in the actual implementation of various smart grid initiatives, among them being demand-side management and the demand response.

With so many stakeholders and potential benefits, the main challenge regarding demand-side management may lie in the development of business cases that properly weigh all the costs and benefits of implementing demand-side management, also considering the possible impact of the market structure and other supporting infrastructure. This is likely where policy and regulation may play a pivotal role, although there are many other factors that may be equally decisive in the development and effective implementation of demand-side management. Some of these are highlighted in the following section.

2.5. Barriers to the Effective Implementation of DSM

A number of factors can act as barriers to the effective implementation of demand-side management, and include technical, structural, regulatory, educational and financial/economical barriers, as further elaborated on in Table 2 [1].

• Financial assistance in the form of loans or direct payments to support investment in energy-efficient technologies
• Direct or free installation of energy-efficient technologies
• Performance contracting, whereby customers get into contracts with utilities to guarantee a certain level of energy performance
• Load control or load shifting, where customers consent to having their energy-consuming devices remotely controlled by the utility in return for financial incentives
• Innovative tariffs, such as interruptible rates, time-of-use rates and real-time pricing, which are meant to improve the levelized cost of electricity supply

The choice of which of these DSM programs to implement in any given context is largely a function of political, socio-economic and regulatory factors, among other considerations. Some DSM programs are geared towards promoting energy efficiency; others target achieving desired load shapes, such as peak-load reduction or load shifting. Broadly speaking, DSM initiatives can be classified into two main categories, based on how load changes are induced [14]:

• Incentive-based DSM schemes are programs facilitated by utilities or energy service providers, which essentially incentivize customers to reduce their energy demand, who then receive compensation for their participation. Reliability conditions or price conditions may trigger the need for load reduction.

• Direct load control (DLC) is a program whereby the system operator remotely controls the on/off state of a customer’s electrical equipment (e.g., space heating, water heater, air conditioner) on short notice. These programs are primarily implemented in the residential and small commercial customer segments

• Electricity market operators may encourage customer active participation, which would lead to more efficient market operation, lower electricity prices and greater innovation in the way of supporting technologies to enhance market operations. Balance-responsible parties may also welcome the opportunity to engage active customers in the balancing markets

• Energy service providers, traders, suppliers and retailers may benefit from providing customer access to the electricity markets in the form of platforms, technologies and products that enable customer participation in electricity markets

3. Supporting Structures for the Effective Implementation of Demand-Side Management

3.1. Enabling Technologies

Demand-side management as a technological evolution can be viewed in the larger context of the smart grid. A smart grid has been defined by the European Commission’s European Technology Platform [16] as: “An electricity network that can intelligently integrate the actions of all users connected to it—generators, consumers and those that do both—in order to efficiently deliver sustainable, economic and secure electricity supplies.” Many other authoritative bodies and organizations in the electric power system sector have given other variants of the definition of the smart grid (e.g., the International Electrotechnical Commission (IEC), the Institute of Electrical and Electronic Engineers (IEEE), the US Department of Energy, etc.) At the core of the smart grid is the use of Information and Communications Technologies (ICT)-based innovative technologies to provide the electric power system with the intelligence necessary to effectively integrate the heterogeneous components (among them, demand-side resources) that need to seamlessly interoperate in order to deliver electricity reliably, securely, affordably, efficiently and sustainably. Some of the key enabling technologies include [17]:

• Information and communication technologies (ICT)
• Grid monitoring and control technologies
• Advanced metering infrastructure and smart meters
• Smart sensor and actuator networks
• Interruptible/curtailable programs offer customers a rate discount or bill credit for agreeing to reduce the energy demand during system contingencies. The customer usually incurs penalties for failures to adhere to the curtailment arrangements. These schemes are typically implemented in large industrial and commercial customer segments

2.2.2. Market-Based Schemes

• Intelligent electronic devices
• Demand bidding enables customers to offer bids to curtail their energy demand based on the wholesale electricity market prices or the equivalent. This scheme is normally open to large industrial customers with a high demand (i.e., with a megawatt (MW) demand threshold specified by the system operator)
• Emergency demand response (DR) participation is offered to customers who can effect load reduction in exchange for agreed-upon incentives during periods of system constraints, such as reserve shortfall, which may endanger the security of supply

3.1.1. Information and Communication Technologies (ICT)

The realization of demand-side management objectives, such as flexible consumption patterns responsive demand behavior and active demand participation in energy markets, requires the evolution of new metering, control and information management technologies that will support the needed functionalities. By availing customers with information such as near-real-time meter readings and real-time pricing data, they can become more aware of the relationship between the level of demand and the cost of electricity supply, on the basis of which they can decide whether to modify their energy consumption patterns or to carry on as normal. ICT technologies facilitate the implementation of DSM schemes in the sense that automated systems can be used to make consumers aware of opportunities to participate in such schemes, and their response or participation can also be automated or carried out manually. The main functionalities required to be provided by the ICT infrastructure for the purpose of implementing DSM include [1]:

• Notification

An electricity market structure refers to the way in which stakeholders in the electric power system interact to produce and deliver electricity to the end customer. The key stakeholders or sectors in the electric power supply chain are generation, transmission, distribution, system operations, wholesale markets and retail supply. Four main market structures can be identified, which are [18]:

• Vertically integrated monopolies
• Measurement
• Unbundled monopolies
  Compliance
• Unbundled electricity market, with limited competition
  Settlement
• Automated controls
• Capacity markets are where customers offer load curtailment as a system capacity to substitute for conventional generation or a lack of transmission capacity or other delivery resources. Prior arrangements are usually made between the system operator and the customer in terms of when the service might be required. Failure on the part of the customer to deliver the service as per the contract usually incurs penalties
• Ancillary service markets are programs where customers bid load curtailments in the ISO/RTO (independent system operator/regional transmission organization) markets as operating reserves. If their bids are accepted, they are paid the market price for committing to be on standby. If their load curtailments are needed, they are called upon by the ISO/RTO to deliver and may be paid the spot market price for the delivered energy

2.2.3. Price-Based Schemes

• Time-of-use (TOU) pricing applies rates with different unit prices for energy consumption at different times. This is usually pre-determined for a 24 h day. TOU rates are meant to reflect the average cost of generating and delivering power during different periods of time
• Real-time pricing (RTP) is a form of rate in which the price of electricity tends to fluctuate at a predetermined time interval (usually hourly), reflecting changes in the wholesale price of electricity at different times of the day. The RTP prices are normally communicated to the customers ahead of time, which could be on a day-ahead or hour-ahead basis
• Critical peak pricing (CPP) rates combine the TOU and RTP rate designs. TOU constitutes the base rate structure, and a provision is made to replace the normal peak price with a much higher price under specified anomalous conditions (e.g., when there is a threat of system reliability being compromised or when supply prices are excessively high)

Some form of notification is required to make consumers aware of opportunities to participate in DSM. These notifications may take many different forms, direct or indirect, passive or active. The medium of delivery of the notifications has to take into account such factors as the volume of notifications that must be delivered (i.e., the number of participants and their geographical spread), the speed at which the notifications have to be delivered and the extent to which customer responsiveness is required (for example, whether it is voluntary or mandatory). Manual notification systems (such as telephone calls or electronic messaging) have traditionally been used for some DSM schemes in the past. These may be suitable when the speed of response is not required to be very fast, and the level of customer participation is low to moderate. In cases where the participant count is very high, however, an automated notification system may be more convenient and more effective. Technological advancements in the way of information and communications technologies are facilitating the implementation of such automated systems, which would otherwise have been technically difficult or economically prohibitive to implement in the past. Automated systems also facilitate improved record-keeping and tracking of how effective the schemes are in engaging customer participation and are especially useful where mandatory compliance may need to be enforced. Measurement of customer participation in (i.e., their contribution to) DSM is important, both for assessing compliance (where there are contractual obligations) and facilitating settlement for the rendered service. The measurement mechanism has to enable differentiating between the “normal” consumption and the DSM-induced consumption modification, on the basis of which the customer has to be compensated. This depends on the design of the DSM scheme, as not all DSM schemes involve direct compensation. In a time-of-use rate or real-time pricing DSM mechanism, for example, it is the customer’s energy use pattern modifications that will directly impact the energy bill, without the customer having to receive direct compensation for participating in the DSM scheme. Measurement can take many forms and will depend on requirements such as volumes of data, the frequency of data transmission, the granularity of data (in terms of time intervals) and the desired speed of response. A means of assessing compliance is needed in order to establish the performance level of the customer and to determine the extent to which they are honoring their contractual obligations, where this is applicable. One methodology used in establishing compliance is referred to as the baseline methodology, which involves estimating customer “normal” consumption and then compensating them on the basis of the variation between the estimated “normal” consumption and the actual consumption (assuming there is a reduction in the actual consumption relative to the one deemed normal). In the case of on-site generation that the customer wishes to bid as a demand-side resource, it is common to use a direct meter that measures the generator’s actual output, which is then used to determine compliance. Where real-time pricing is used as the DSM mechanism, compliance is not necessarily required to be determined, since this is a more direct mechanism for inducing customer energy consumption modification, and the customer is simply charged the hourly energy price based on their magnitude of consumption. A settlement system is required, which acts as a means to credit the customers participating in the DSM schemes. The system provides facilities for the maintenance of meter usage, market pricing, event compliance levels and individual contract terms. Large consumers who have direct access to the wholesale electricity markets receive a direct settlement from the wholesale market. For the majority of DSM participants, however, their participation may be facilitated by such entities as aggregators, distribution companies, or energy service providers, among others. Settlement for the rendered service in that case is a two-step process: from the wholesale market to the DSM scheme facilitator (e.g., aggregator), and then from the DSM facilitator to the participating consumer. The settlement system design has to provide for delineating between these stages of settlement. Automated controls enable DSM schemes where the utility operator or other service provider is able to remotely and automatically control predetermined loads. This is commonly applied in the residential demand sector. The loads that are typically controlled this way include heating, ventilation and air conditioning (HVAC), electric water heaters, pool pumps and lighting. Automated controls may also be found to a somewhat lesser extent in the commercial sector, with the typical remote load control technology being lighting control. Building automation control technologies exist, which enable systems to be programmed so as to respond to electricity price signals rather than simply demand levels. Thus, they are able to adjust the energy consumption levels at times when the predetermined settings indicate that electricity prices are in excess of predefined thresholds.

3.1.2. Grid Monitoring and Control Technologies

A modern integrated power system requires high-level surveillance, monitoring and control in order to maintain the desired level of security, reliability, efficiency and quality of power supply. The main components of the modern power system include the bulk generation, transmission and distribution systems, as well as electricity markets, operations, utilities and end-consumers. Grid monitoring and control requirements are differentiated by the part of the system of interest. The transmission system, as the backbone of the entire interconnected power system, for example, requires advanced technological tools for various analytics such as real-time stability assessment, robust state estimation, dynamic optimal power flow, contingency analysis and security assessment. Key technologies employed in these analyses include phasor measurement units (PMUs), state estimators, advanced metering infrastructure and smart meters, smart sensors and actuator networks and intelligent electronic devices (IEDs). The distribution system acts as the interface for industrial, commercial and residential consumers of the bulk electric power supply system. The main monitoring and control needs at this level include smart metering, automated meter reading and automatic billing, fault detection, isolation and service restoration, feeder reconfiguration, voltage optimization and demand-side management. The key objective in the implementation of advanced technological and analytical tools in the entire electric power supply chain is to realize a self-monitoring, self-healing and resilient smart grid network that is self-aware and is capable of taking actions independently on the basis of situational awareness.

3.2. Market Structures

• Unbundled electricity market, with full competition
• Extreme day pricing (EDP) is similar to CPP (in having a much higher price than normal) and differs in the sense that EDP is in effect for the whole 24 h of the extreme day and is unknown ahead of time [15].

2.3. Drivers for DSM

The primary driver for demand-side management is to enlist the support of customers in optimizing electric energy production costs and enhancing energy utilization, which in turn contributes to energy security and reliability. There are, however, many other factors that may act as drivers for the implementation of demand-side management, some of which are listed in Table 1 [1]: The vertically integrated monopoly structure is the traditional system where the electric utility controls and undertakes all of the functionalities of the electric power supply system, including generation, transmission and distribution. The structure does not have an open electricity market, meaning there is no competition in the delivery of the various electricity supply-related services. An unbundled monopoly is similar to a vertically integrated monopoly, the main difference being the separation of generation from all other functions of electricity supply (i.e., transmission, distribution and wholesale and retail markets). This structure also does not encourage competition, except at the generation level, where large generators can compete to supply to distribution companies, and perhaps large industrial consumers as well. An unbundled electricity market with limited competition introduces a competitive wholesale electricity market, where many different generation companies bid to supply to electricity distributors, and possibly large industrial customers as well. In this structure, generation companies have open access to the transmission and distribution systems, and there is competition for the supply of generation at the marginal cost of supply. Depending on the size, the wholesale market may also be open to self-generating large consumers and independent power producers. In an unbundled electricity market with full competition, all the major sectors of the electricity supply system (i.e., generation, transmission and distribution) are separated, and there is competition at all levels, that is, wholesale and retail markets. This structure is considered to be the most advanced in terms of embodying the objectives of modern deregulated power systems with open access to electricity markets [1]. The market structures described above can be distinguished by the degree of unbundling and the extent of competition in the wholesale and retail electricity markets. In terms of the implications for demand-side management, the unbundled electricity market with full competition is the one that is most likely to fully exploit the potential of demand-side management to contribute to efficient network operation. Power systems in developed countries such as the U.S.A. and European countries have largely reached this highest level of unbundling and competitiveness of electricity markets.

3.3. Policies and Regulation

One of the major issues or challenges in the development and implementation of DSM is the implementation of various policies and regulations. Establishing a unified standard policy system that addresses various aspects of DSM development has been a challenge because of the variation noticeable in regions. The imbalance in the regional development of DSM can be attributed to different key factors. First, there are significant disparities in economic development and energy utilization across different regions, leading to variations in the progress and implementation of DSM initiatives [19]. Second, there are notable variations in the development of DSM across different industries. DSM encompasses multiple sectors, with heavy industry being a major consumer of power and a key focus area for DSM implementation [20]. Due to the regional imbalances in DSM development, it is challenging to establish a unified DSM mechanism and formulate standardized policies. Power grid companies often face resistance from local governments when implementing DSM due to the lack of a cohesive approach. Addressing the imbalance in short-term development is difficult, necessitating a long-term vision. Consequently, it is crucial to establish a long-term DSM policy mechanism that gradually promotes the balanced development of DSM across regions. These policies create a favorable environment for DSM development and play a significant role in facilitating and guiding the implementation process. Countries such as China have formulated new policies to support the development of DSM. One such regulation was the Power Demand Side Management Regulation, which provides clarity on the scope of work and identifies the primary implementing body and liability subject for Demand-Side Management (DSM). It also addresses organizational management, technical measures, power pricing, funding sources and other relevant aspects. The regulation’s provisions ensure the smooth implementation of DSM initiatives [20].

3.4. Demand-Side Resources

In principle, demand-side management is aimed at somehow shaping the load profile in a way that leads to an improvement in the technical and economic operation of the power system and increases the security and reliability of supply. This is mainly achieved by inducing energy consumers to modify their energy consumption behavior so that the load can “follow” the generation to some extent, which is the opposite of the traditional approach to supply–demand balance, where the generation has to follow the load. Demand-side resources encompass more than just energy-consuming devices and equipment. With the significant progress made in the integration of distributed energy resources into the distribution systems, these are generally considered to form part of demand-side resources, to the extent that they are largely not under the direct control of the utility operator. Distributed energy resources (DERs) comprise conventional and non-conventional distributed generation, renewable energy sources (such as photovoltaic and wind) as well as energy storage [21]. DERs are anticipated to have an increasingly large contribution to distribution system operation. The high penetration of distributed generation—and especially variable renewable generation—in the distribution network is another factor that increases the need for flexibility resources in the network. This is mainly due to the fact that the DERs are not under the direct control of the utility operator, and thus they are considered to be non-dispatchable in the traditional sense of dispatchability. In other words, their behavior is to a large extent unpredictable, at least from the perspective of the utility operator. Moreover, for variable renewable generation, such as solar and wind generation, the output is intermittent and cannot be accurately predicted ahead of time. Thus, by implementing effective demand-side management and demand responses, the potentially adverse impacts of the DERs on network operation can be partially alleviated, in addition to the main objective of load shaping and load levelling that demand-side management is intended to accomplish [3]. This in turn can enable the integration of more variable clean energy into the electric grid, which is one of the main objectives of modernizing the electric power system, leading to higher energy security and an environmentally sustainable energy supply [9].