2. Smart Grids: Motivations, Goals and Paradigms

The essential aim of grid management is to match supply and demand. SG technologies help to achieve this goal by allowing control from the demand side (which is impossible to attain with conventional/classical control models, which rely exclusively on supply modulation). SGs incorporate the more recent advances in information and communication technologies (ICT) to gather and consolidate the relevant information from all the players in the energy chain—from producer to consumer—for accomplishing the following objectives ^[6][35][36][6,63,64]:

• Better integration and more economical, holistic, and rational management of distributed production (i.e., including distributed micro-generation resources);
• Curb GHG emissions by increasing the penetration of RES, improving their use, and minimizing production waste;
• Overall improvement of the quality of the energy supply (i.e., better voltage and frequency stability, better reliability, and consistency);
• Development and incorporation of demand response and demand-side resources, enhancing overall energy efficiency (essentially by reducing losses due to overloading of lines and transformers) and avoiding the need for fossil-fueled backup facilities;
• Enable the active participation of all consumers in the vital load balancing function (including small residential consumers);
• Implementation of real-time automated technologies that optimize the operation of appliances and consumer devices;
• Fully exploit the use of the latest information and control technologies to improve overall reliability, security, and efficiency of the energy infrastructure;
• Integration of advanced electricity storage and peak-shaving technologies, including plug-in and hybrid EVs, and residential thermal energy storage (e.g., HVAC and refrigerators/freezers);
• Better interoperability between transport and distribution networks (thanks to enhancements in communications and automation).

2.1. The Urgency for Smarter Grids

All the players in the energy sector—producers, transport and distribution owners, managers, governmental and non-governmental regulators, consumers, and their associations—consensually recognize the urgent need of acting on several challenging problems, affecting the present and future of the global electrical infrastructure ^[37][65]. Some of the main motivations leading toward the technological advancement of electrical grids are [6]:

• Intermittency and fluctuations due to wind and solar generation: Wind and solar production are characterized by large (often unpredictable) fluctuations ^[38][66]. This caps the share of these RES, since there is a need for extra controllable and rapid response generation capacity (i.e., thermal and hydro, usually), and/or energy storage facilities (e.g., pumped hydro, electrochemical batteries, thermal phase-change, flywheels, etc.), and/or DSM or peak load shifting ^[39][40][67,68], to assure the required stabilization ^[41][42][43][69,70,71].
• Climate change: The need to respect the agreed emission reductions (the Paris Agreement and global political pressures towards the reduction of CO₂ emissions) is pushing the investment in renewables, with their associated problems for the current infrastructures ^[44][72]. In this context, the implementation of SG paradigms becomes indispensable.
• Rising energy prices: The short, medium and long-term rising trend of oil and gas prices (due to diminishing reserves, conflicts, politics, instabilities, and market fluctuations) are among the strongest motivations supporting the increasing penetration of RES and efforts to enhance energy efficiency ^[45][73^[46],74]. The recent invasion of Ukraine by Russia has caused an unprecedented spike in oil prices, which introduced additional instability and unpredictability in oil and gas prices.
• Ageing, degradation, and obsolescence of infrastructures: Capacity limitations, high losses, lack of reliability, and high maintenance costs are some of the problems associated with old infrastructures (some of them date back to the 1950s and are still in full use); the adoption of SG paradigms will optimize the usage of the existing resources, thus curbing the upgrade costs and allowing more time for updating and/or replacing the degraded infrastructures ^[47][75].
• Increase in global energy demand: The existing infrastructure is being overloaded by the increase in energy demand, particularly in developing countries; this leads to a lack of reliability, quality, and consistency of the supply. Despite the significant drop in 2020, demand is still about 9% higher than in 2010 and is projected to continue rising ^[14][48][49][26,76,77]. Without the adoption of SG management models—actively involving the consumer—it will be very difficult (or even impossible) to respond adequately to this growth by investing solely in physical infrastructures.
• Inefficiencies resulting from the need for backup generation capacity: The large demand fluctuations, and the lack of synchronism between the demand and renewable production, lead to the underuse of the installed generation capacity, resulting in a considerable financial burden (i.e., the backup generators are inactive most of the time, being only used in very limited peak periods) ^[34][12].
• Increasing distances between production, storage, and consumption sites: RES production centers (i.e., hydro, wind, and solar) are often placed far away from the population and industries, requiring long transport infrastructures. This diverts financial resources that would be, otherwise, applied to the maintenance, renovation, and upgrade of current facilities. This also increases transport and distribution losses ^[50][78]. Distributed production and peer-to-peer (P2P) energy trading can attenuate these issues, but can also exacerbate the challenges for efficient grid control ^[50][51][52][78,79,80].
• The growing profusion of small independent energy producers: Small production installations (i.e., mini/micro-generation) of solar photovoltaic, wind, hydro, biomass, co-generation, etc., increase the difficulties in balancing the grid from the supply side, increasing the need for innovative communication and distributed control models and technologies ^[51][52][79,80].
• New energy consumption patterns: Energy consumption patterns have been changing considerably in the last decade, particularly in the residential sector ^[53][81]. The increasing profusion of electrical plug-in vehicles (EVs) ^[54][82], smart homes, and smart buildings offers new opportunities for control from the demand side, including load control and energy storage—e.g., charging EVs and home battery banks exclusively in off-peak hours, storing latent heat in HVAC systems, etc. ^[55][56][57][83,84,85].
• Energy market liberalization: The liberalization of the energy markets in developed countries increased the need for energy price transparency ^[58][59][86,87]. This has led to the adoption of technologies such as smart metering ^[60][88], which opens additional opportunities for implementing smarter models for DSLM by facilitating the collection of relevant consumption information and—in some more sophisticated implementations—allowing bidirectional communication.
• Expanding transnational grid interconnections: To best balance supply and demand—particularly with the increasing penetration of RES—transnational grid interconnections are becoming increasingly important ^{[61][62][63][64]}[89,90,91,92], introducing additional control challenges and communication issues ^{[65][66][67][68]}[93,94,95,96].

2.2. Increasing the Penetration of RES: Challenges to Grid Control

Matching production and demand is critical for the stable operation of electrical grids. Unfortunately, wind and solar generation are infamous for their lack of consistency: they cause significant fluctuations in production, which are frequently out of phase with the demand ^[6][69][6,97]. This could, theoretically, be solved by implementing enough storage capacity. Regrettably, large to medium-scale storage solutions (e.g., pumped hydro, thermal storage, and lithium battery banks) are very expensive and/or complex to implement ^[70][71][60,98]. Consequently, they currently still have a relative residual effect ^[72][99]. This results in the necessity of keeping fossil fuel-fed power plants, for the sole purpose of stabilizing and regularizing the grid voltage ^[73][100]. In the limit situation, the installed capacity of fossil fuel-fed generators needs to be matched to an equal capacity of wind and solar production, resulting in significant economic and environmental impacts. Moreover, during peak production and low demand, most of the energy produced by these RES is not even injected into the grid, leading to significant underuse of their capacities ^[69][97]. Balancing production with demand is most efficiently performed from the demand side since—among other advantages—this approach lessens the burden on the transport and distribution infrastructure. Thus, considering the inevitable increase of RES in the energy mix, the active participation of the smaller domestic consumers in the regularization of the grids is increasingly being seriously pondered ^[74][75][101,102].

2.3. Demand-Side Load Management Paradigms

The expression “Demand-Side Management” was coined, in the early 1980s, by the Electric Power Research Institute (EPRI), being defined as (paraphrasing): “planning, implementation, and monitoring of electrical networks with the objective of influencing the consumer to use electricity in a manner that produces desirable changes in the network load profile, i.e., reducing the magnitude of its fluctuations” ^[76][77][78][103,104,105]. This comprises all the strategies and models aiming at influencing the consumption profile in the direction of transferring load from peak to off-peak periods ^[39][78][67,105]. The greatest virtue of this approach resides in the fact that it is significantly less costly to manage the load profile from the consumer side than to build a new power plant and/or increase the carrying capacity of the transport and distribution infrastructure ^[77][78][79][104,105,106]. DSLM is especially decisive in the context of increasing energy production from RES since it can provide a means of mitigating the supply and demand imbalances caused by the inherent fluctuating and unpredictable character of some of these energy sources. Particularly, solar and wind intermittency constitutes the main obstacle to the growth of their share in the energy market ^[7][80][7,107]. Since the fundamental actors in DSLM are the consumers, these need sufficient motivation and incentive to incur the inevitable costs that the installation and operation of DSLM functionalities might imply, i.e., some financial advantage needs to be offered to the energy customers by the energy providers.

2.3.1. Price-Based Demand Response (PBDR)

One way of achieving a more rational load profile is by implementing price models that reflect the relation between demand and supply, thus incentivizing the implementation of load management (LM) methods and technologies by consumers ^[81][82][108,109]. This is the most common approach adopted by energy suppliers to motivate consumers to modify their consumption patterns, in a way that helps balance the grid, by attenuating demand peaks and, thus, minimizing costly interventions from the supply side ^[83][84][85][110,111,112].

• Time-of-use pricing(ToU): This is the currently most practiced variable price regime in the majority of countries since it is the most straightforward to implement: it defines differentiated rates for two or three daily periods, e.g., peak, mid-peak and off-peak. It is relatively trivial to implement a cooling equipment control firmware for managing the consumption in the context of this type of pricing schedule, minimizing the energy bill ^[86][87][88][113,114,115].
• Critical peak pricing: According to this pricing scheme, a normal rate (usually belonging to the previously described ToU family) is valid most of the time, during the year ^[83][110]. However, during known occasions with exceptionally high demand (or, conversely, low supply), a higher rate is applied. These periods happen only for a few days, or even a few hours, during the year ^[83][110]. For the consumer, the advantage of this scheme is that the regular rate can be kept lower than it would be possible otherwise. However, that is not, ultimately, the goal: the main objective is to curb the demand by motivating the client to consume less during the short critical high-demand/low-supply periods ^[88][115].
• Real-time pricing: In this pricing system, the energy supplier broadcasts to the consumers, continually, the constantly varying energy rate. This quotation indirectly represents the continually updated relation between supply and demand, as created by the market dynamics. Energy producers dynamically increase the energy price during high-demand/low-supply periods, discouraging consumption during these critical periods ^[88][89][115,116]. In an SG framework, an automatism can, conceivably, be implemented to react instantly to continuously updated price fluctuations, using all available controllable resources (i.e., manageable loads and energy storage devices) to reduce the consumer energy bill and, thus, indirectly help to balance supply and demand ^[74]101^[83][,110^88][,115⁸⁹,116,117]. Despite their designation, current implemented plans are not truly/strictly “real-time”. There are, presently, two main modes of so-called “real-time” pricing being practiced: one broadcasts the quotations a day in advance (DA-RTP) and the other gives the hourly price within 60 min in advance (RT-RTP).

2.3.2. Incentive-Based Demand Response (IBDR)

This class of client-side energy demand modulation is based on incentives that are not directly connected to variable prices or quotations ^[91][92][118,119]. A contract is established between supplier and client, according to which the latter promises to respond to supplier demands for load restraints, depending on contingencies. In turn, the supplier awards the client with financial incentives (e.g., lower energy rates and/or one-off discounts) ^[91][92][118,119].

2.3.3. Direct Load Control (DLC)

In this system of demand control, the energy supplier directly controls the loads from the client side. The energy supplier benefits from an augmented ability to balance supply and demand, and the client receives financial rewards (e.g., lower rates and/or discounts) ^[83][110]. For larger clients, the supplier usually installs the necessary telemetry and telecontrol hardware at the client’s site ^[88][93][94][115,120,121]. For domestic consumers, it is possible to implement direct control of appliances by factory-equipping them with electronic control units (ECUs) with appropriate firmware and Internet connectivity. Validation can be achieved with an Advanced Metering Infrastructure (AMI) (i.e., “smart meters”). Currently, programmable communicating thermostats (PCTs) have been successfully implemented in some regions.

2.3.4. Demand Bidding and Buyback

In these programs, a sophisticated automated negotiation protocol is established between energy suppliers and consumers. They both agree on the price—on a case-by-case basis and according to market fluctuations—for a predetermined quantity (or package) of voluntary cuts from peak-load demand made available by the consumer to the supplier ^[92][95][96][119,122,123]. These programs are currently being tried, with some success, with large consumers (e.g., industry, large commercial and office buildings). There is no fundamental obstacle to implementing these schemes for small residential consumers, although it is certainly more complicated and might prove to be, ultimately, impractical.

2.3.5. Emergency Demand Response (EDR)

In extraordinary cases of insufficient production and/or excess demand, energy producers set up a contingency program that includes communication with clients who can cut their load immediately, i.e., clients with energy storage facilities, backup generators, or the ability to postpone or throttle loads down—e.g., water heaters, HVAC, etc. ^[97][124]. The success of this approach depends on the ability of the supplier to predict such critical periods with enough antecedence, and on the existence of an open and effective communication channel between suppliers and consumers. This channel should be ideally telematic/digital, and the response from the demand side should be also automated, e.g., a group of electrical loads is deemed non-essential and is shut down during critical periods. Alternatively, variable power loads can be throttled down ^[98][125]. The fundamental difference between this modality and DLC is that the consumer is the one responsible for controlling the loads and not the supplier. Like with other programs, the client is rewarded with financial benefits ^[88][115].

2.3.6. Distributed Frequency Regulation Services

Consumers who adhere to this type of program help the suppliers maintain the stability of the AC frequency. Clients allow the supplier to directly modulate their loads, reacting quickly—in real-time—to frequency fluctuations ^[99][126]. This implies the implementation of DLC, i.e., by installing specialized equipment on the client’s site, for frequency monitoring, telemetry, and load control. Again, the client receives incentives or financial compensation ^[83][91][92][110,118,119].

2.4. Practical Models of Load Management for the Household

In practice, the involvement of domestic consumers in DSLM can be classified in three levels, in increasing order of SG integration:

• ToU—Load shifting based on fixed time-of-use pricing differentiation: the consumer adheres to a pricing scheme where the supplier defines two or three fixed daily periods with differentiated tariffs—e.g., peak, mid-peak, and off-peak (these periods can change according to the season). Based on this, the consumer schedules the consumption to minimize the energy bill. “Smart” appliances (including refrigerator-freezers) can automatically (on a time-programmed basis) decide to postpone consumption to off-peak periods, whenever possible. In this “low-level” scheme, there is no need for continuous communication between consumers and suppliers. Thus, this scheme might not be considered entirely within the realm of an SG (it might be classified as “SG level zero”). However, it still implies the use of smart meters that can record the consumption that occurs within the two or three defined periods; this approach implies no costs from the energy provider side: the consumer can simply configure the time-programming control of the appliance with the ToU periods;
• RTP—Load shifting based on “real-time” forecasted pricing: instead of fixed daily periods, the supplier quotes the energy price to the consumer 24h to 1h in advance (usually an hourly price, but it can also be in fractions of an hour). The supplier can forecast future supply and demand based on various predictable factors (e.g., weather forecasts, which determine RES production). Since it is humanly impractical for the typical domestic consumer to manually manage the load optimally in this scheme (particularly for short in-advance periods), the process should be automated via “smart appliances” with IoT connectivity, allowing these to continually receive the real-time pricing from the energy supplier and algorithmically decide when to postpone consumption; costs from the provider side are relatively negligible: it only needs to broadcast the price information via the Internet and charge the client accordingly, based on the time-referenced consumption recorded by the smart meter (assuming one is already installed);
• DLC—Load management based on direct appliance control by the energy supplier: in this high-level SG scheme, the consumer grants the energy supplier direct control of some “smart appliances”. The simplest implementation of DLC for the domestic consumer is the “remotely controllable HVAC thermostat” (PCT). The extension of DLC to other appliances (including refrigerator-freezers) implies the implementation of more sophisticated bidirectional IoT protocols (which implies an initial investment in information systems by the energy provider, although of relatively low financial impact). The energy supplier can use this facility to aid in load balancing, frequency regulation and emergency demand response (EDR). To make this economically advantageous for the consumer, the energy supplier can offer attractive discounts and pricing, allowing the consumer to reach a short break-even period on the required initial investment.
• ^][90][

2.5. Smart Grid Enabling Technologies

A host of new technologies enables the participation of small residential consumers in SGs: Internet-of-things (IoT), smart meters (SMs) and advanced metering infrastructures (AMI) ^{[60][100][101]}[88,127,128], home automation (HA), home area networks (HAN), neighborhood area networks (NAN), wide area networks (WAN) ^[102][129], meter data management systems (MDMS), home energy management systems (HEMS) ^[74][85][101,112], and building energy management systems (BEMS), just to mention the more relevant ones ^{[83][103][104]}[110,130,131]. These technologies provide a workable environment for DSLM/DR programs, harmonizing supply and demand via the modulation of consumption. The participating consumers receive financial incentives from the energy suppliers, rewarding their participation in these programs ^[83][91][92][110,118,119]. In a Communication-Based Demand Response (CBDR) program, household appliances communicate with SMs, establishing an automated bidirectional communication between domestic consumers and energy suppliers. Consumer appliances receive updated information about energy prices—in real-time or not—and can, autonomously, make load management decisions, using their capacity to delay consumption and/or store energy (thus minimizing the electricity bill). Reciprocally, appliances can offer their available capacity for on-demand immediate consumption to the energy marketplace, where they can be commanded by suppliers ^[105][132].

2.5.1. Advanced Metering Infrastructures (AMI)

One of the assets that eases the evolution toward smarter electric grids, is the growing installed network of SMs in developed countries. These apparatuses contain a microcontroller, memory, updatable firmware, and means of remote communications with the energy suppliers (generally via PLC and/or GSM) ^{[100][101][106]}[127,128,133]. The American Federal Energy Regulatory Commission (FERC) defines Advanced Metering Infrastructure as “meters that measure and record consumption data, hourly or with greater frequency, supplying this information to both users/clients and energy suppliers, at least daily, that information being used for billing and other purposes” ^[83][110]. SMs are essential for the implementation of SGs ^[60][107][88,134]. Their most basic function is to record daily consumption history, allowing ToU billing. Nevertheless, they can also offer several other functions in the context of DSLM/CBDR ^[75][108][102,135]. For example, these apparatuses can serve as a communications gateway with electrical loads that can be remotely controlled. They can also supply vital real-time information about where consumption is taking place. Importantly, they can also validate both energy saving and/or load shifting actions by the consumers ^[109][136].

2.5.2. Wide Area Network Infrastructures

A fundamental requirement for the implementation of SG functionalities is the availability of communication means between all the players (i.e., production, transport, control, distribution, and consumption). Fortunately, nowadays, wide-area wireless Internet access technologies are globally widespread. Güngör et al. (2011), Gao et al. (2012), Usman and Shami (2013), Supriya et al. (2015), Mahmood et al. (2015), and Emmanuel and Rayudu (2016), conducted very comprehensive reviews of available communication technologies and standards for SGs ^{[103][104][110][111][112]}[130,131,137,138,139], which are still broadly applicable and up to date.

• Mobile cellular infrastructure: The widespread coverage of cellular infrastructure is a valuable resource for supplying the necessary means of communication for the implementation of SG paradigms. Given the relatively modest bandwidth needs of these applications, even the most basic GSM/2G coverage can be sufficient for this purpose ^{[103][104][110]}[130,131,137]. Many SMs are equipped with GSM connectivity.
• Power Line Communication (PLC): Since all the constituent parts of the electrical grid are, inherently, physically interconnected by conductor wires, it is logical to use this pre-existing infrastructure as a communication medium. This would have the advantages of not depending on third-party infrastructures and allowing an economic means of communication between all grid constituents. Frequently, SMs are equipped with PLC transceivers. Plentiful technologies, protocols, and standards exist for this purpose—e.g., IEC 14908-3 (Lon Works), IEC 14543-3-5 (KNX, BUS), CEA-600.31 (CEBus), IEC 61334-3-1 (DLMS), IEC 61334-5-1, IEEE 1901.2, ITU-T G.henm, PRIME, G3-PLC, IEEE 1901, TIA-1113 (HomePlug 1.0), ITU-T G.hn (G.9960/G.9961), and HD-PLC ^[103][104][130,131]—but their implementation is still relatively limited and of local/regional scale.

2.5.3. Internet-of-Things

According to the Recommendation, ITU-T Y.2060 of the Telecommunication Standardization Sector of the International Telecommunications Union (ITU), the IoT “(…) can be viewed as a global infrastructure for the information society, enabling advanced services by interconnecting (physical and virtual) things based on existing and evolving interoperable information and communication technologies (ICT)” ^[113][140]. It is unviable to list all the protocols encompassed under this concept: they are numerous and still growing (e.g., ZigBee ^[114][141]). The relevance of this definition in the current context is justified by the fact that the exchange of information between SG devices will adopt some of these protocols ^{[24][80][115]}[36,107,142]. Furthermore, the application of IoT technologies for real-time demand response in the household, specifically narrowband-IoT (NB-IoT), is presently being investigated ^[116][143], particularly in the context of the newest SG paradigms ^[117][118][144,145].