Demand Response Implementation: Overview of Europe and United States Status

Abstract

The authors review the efforts made in the last five years to implement Demand Response (DR) programs, considering and studying several models and countries. As motivation, climate change has been a topic widely discussed in the last decades, namely in the power and energy sectors. Therefore, it is crucial to substitute non-renewable fuels with more environment-friendly solutions. Enabling Distributed Generation (DG), namely using renewable resources such as wind and solar, can be part of the solution to reduce the greenhouse effects. However, their unpredictable behavior might result in several problems for network management. Therefore, the consumer should become more flexible towards this new paradigm where the generation no longer follows the demand requests. With this, Demand Response (DR) concept is created as part of this solution. This paper studies the European Union and United States’ current status, with over 50 references.

1. Introduction

To meet the energy-related points of the Sustainable Development Goals (SDGs), the energy market members must focus on innovative markets for flexibility and ancillary services [1,2,3,4,5]. For instance, new business models must be implemented in existing markets considering the aggregation of active players [5]. Some key regions are making efforts [2]. In 2018, China was approaching full deployment to all electrified endpoints. Japan, Spain, and France were preparing to introduce the paradigm, considering that smart meters reached around half of the wholesale market [3]. Besides, that year’s Demand Response (DR) capacity increased by 4%, keeping the average growth rate from the past five years regardless of the potential and progress made in the smart-meter deployment in several core regions. However, according to Lin et al. [3], the problem still needs to be solved by developing market designs and business models that can extract the full capability from the available flexibility.

1.1. Contextualization and Background

The electricity demand from consumers in power and energy systems was considered rigid in the prior paradigm. Introducing the Smart Grid concept, focusing on a consumer-centric approach, might have countless advantages, namely flexibility markets [3,4,6,7,8]. The main players are [6]: Transmission System Operators (TSO), Distribution System Operators (DSO), Balance Responsible Parties (BRPs), aggregators, and retailers. The distribution system’s operation is the responsibility of the DSO, while the transmission system’s service and stability are the responsibility of the TSO [7]. Collaboration between TSOs and DSOs is essential for realizing the flexibility’s full potential [8]. A market entity, such as a wholesale supplier or retailer, or its designated representative can perform the BRP function, which entails resolving the imbalance and paying penalties for deviations from energy schedules [4]. Lastly, the retailer is an established business providing electrical energy to consumers.

The market model used determines the role of an aggregator. The aggregator gathers flexibility through active consumers and renewable-based resources. It can operate in different network parts using the connected resources to deliver electricity and ancillary services. This entity can also be an intermediary between small players and the wholesale market [9,10,11,12]. The aggregator can also gather flexibility through renewable-based and active consumers (through DR) acting as a Local Energy Community (LEC) manager. In [10], LEC is defined as “an association, a cooperative, a partnership, a non-profit organization or other legal entity which is effectively controlled by local shareholders or members, generally, value- rather than profit-driven, involved in distributed generation and performing activities of a distribution system operator, supplier or aggregator at the local level, including across borders.” LEC sends different signals for the active players to participate in different contexts to help balance the grid. With this, each consumer can use a set of appliances, which do not have a fixed schedule, to provide flexibility in wholesale and ancillary power markets—throughout DR. Numerous DR definitions have been proposed in recent years. A common definition reads [11]: “…tariff or program … to motivate changes in electric use by end-use customers … changes in the price of electricity over time, … incentive payments … high market prices… grid reliability…”. A more recent one, published in European Directive 2019/944, says [10]: “… change of electricity load by final customers… market signals… time-variable electricity prices or incentive payments, … final customer’s bid to sell demand reduction or increase… market… alone or through aggregation”. There are two main options for these programs: price-based (also known as implicit demand response), which uses price signals and tariffs to encourage DR event participation, and incentive-based (also known as explicit demand response), which compensate players for the flexibility they receive through direct payments [12].

Figure 1 provides a detailed visualization of the procedure from the announcement—the deployment.

The DR program manager (possibly an Independent System Operator (ISO) and Regional Transmission Organization (RTO)) must notify the Aggregator of the DR event until the announcement deadline (δ), starting the next phase thereon [14]. The ramp period is when consumers must reach the contractual DR event baseline. The aggregator faces a more difficult task due to the variable duration of the advance notification to players, which can last anywhere from several months to five minutes before the ramp period [15]. The aggregator and DR program manager exchange information and setpoints in an upper communication layer, specifically the to-be-implemented DR event, the assessment duration, and the ramp period [13]. Following the timeline and during the ramp period, the communication in the lower layer is conducted by a cascade process, considering different levels and iterations until reaching a goal point equal to or above the reduction baseline [13]. The Deployment period is the moment during the ramp period when the Aggregator takes the initiative to begin the event—it could be later than the ramp period starts. All the information regarding DR amounts is collected [16]. To a better interpretation of these phases, focus on TDR2, where the forecasted reduction baseline (φ) is not yet achieved, and two different intermediated periods are considered: activation notification period (αDR2) and actual response period (βDR2). The aggregator notifies the consumers regarding the difference between the actual and the goal value—αDR2. The consumers must reply, and those with positive answers will start a load reduction process demonstrated in βDR2. However, although a manager and the active consumer must have a contract, there is always the possibility of non-response [17]. Still, once the forecasted reduction baseline is achieved, the DR program manager must be informed. In the case of insufficient reduction in the actual period to start the DR event, the aggregator can implement a new TDRN until the reduction deadline moment (θ). However, the DR program manager defines a margin of forecast error (∆E). The sustained period can begin if the available reduction capacity exceeds the reduction baseline (σ). The actual DR event starts at this point, and the participants must maintain their committed level of reduction.

1.2. Motivation and Contributions

DR programs (DRPs) were first implemented in the United States (US) in 1970, as Mohammad Shakeri et al. [18] recall, to control peak hours. So, the US experience gives them the lead for the implementation of DR but only for a little since the European Union (EU) countries are taking important steps to engage the demand side resources in the real systems [19]. According to Net Zero Emissions (NZE), by 2050, the flexibility of the electricity system, which is required to strike a balance between wind and solar and shifting patterns of demand, will quadruple, even as retirements of fossil fuel capacity will reduce conventional sources of flexibility [20]. All forms of flexibility must be significantly increased due to the expected energy transition. Energy Storage Systems (ESS), DR, and flexible, low-carbon power plants supported by electricity networks should be smarter and integrated into a successful business model. Furthermore, it is necessary to improve electricity systems’ resistance to cyberattacks and other emerging threats due to potential privacy issues [12].

Bearing in mind these facts, the authors’ goal for the literature review is to resolve the following research question:

What are the current developments, lessons learned, and future perspectives for DR in the US and Europe?

It is known that in the literature, there are already several reviews and works on the DR topic [21,22,23,24]. Still, none of the recent works assess US and EU markets, analyzing their strategies to implement DR, future perspectives, and plans.

Table 1. Main discussed topics and previous works.

Ref.	DR Players	Europe	United States	Network Management	DR Initiatives	DR Market Analysis
[21]	x			x	x
[22]		x		x		x
[23]			x		x	x
[24]	x	x		x	x
[3]	x	x			x	x
[25]				x	x
This work	x	x	x	x	x	x

shows the main discussed topics and compares the selected works and the work developed by the authors. It must be highlighted that distributed energy resources do not consider only DG, but also mention ESS and Electric Vehicles (EV) within their study.

The reviewed literature covers various topics, such as integrated demand response, energy transition, DR underinvestment, and scheduling problems in virtual power plants, with various recommendations provided to address these issues [3,21,22,23,24,25].

Starting with Malik Ali Judge et al. [25], their study surveys various smart grid frameworks and their social, economic, and environmental impacts from 2015 to 2021. It explores the integration of renewable energy sources, energy storage systems, and plug-in electric vehicles, which provide long-term economic and environmental benefits, and guidance for researchers and transmission and distribution system operators. Besides, Daiva Stanelyte et al. [3] discuss the need for the electricity sector to increase transmission capacity through technological and social modifications. DR policy proposals emphasize competitive pricing and consumer participation in reducing carbon emissions and implementing smart energy solutions. Their work explores the latest methods of DR implementation within Europe countries, considering other sectors for future work. Moreover, Huang et al. [21] analyzed the integrated DR. In this concept, all energy customers can participate in DR by developing a multi-energy system that combines electricity, heat, natural gas, and other forms of energy.

Furthermore, Karlo Hainsh et al. [22] focused on energy transition, comparing many scenarios studies to find similarities in the pathway results and expand on the consolidated findings relevant to the EU Green Deal. According to the findings, high electrification rates are imminent for rapid decarbonization. However, to accelerate the energy transition, strong policy enforcement in the short term must go hand in hand with technology development and deployment.

From the US side, Marilyn Brown and Oliver Chapman [23] analyze the inexplicable DR underinvestment in Georgia. These authors’ findings suggest that the need for financing initiatives, market innovations, infrastructure modernization, and enablers of socio-economic inclusion are more to blame for the DR gap than technology limitations. Finally, Hossein Rouzbahani et al. [24] reviewed the scheduling problem in virtual power plants. Questions related to frequently used scheduling techniques, considerations of technical and economic aspects, and dealing with different types of uncertainties were addressed. The paper suggests utilizing a Deep Reinforcement Learning (DRL)-based technique to address these concerns due to generalization, scalability, and feature extraction.

The first two studies brought a broader view of DR within the real market. Moreover, Huang et al. [21] and Karlo Hainsh et al. [22] emphasize the importance of incorporating multiple energy systems and high electrification rates to achieve a sustainable energy transition. Marilyn Brown and Oliver Chapman [23] also highlight the need for financing and market innovations to drive demand response investment. In contrast, Hossein Rouzbahani et al. [24] focus on the technical challenges of scheduling in virtual power plants and suggest using DRL-based techniques to address these issues. Still, the review performed within the scope of this paper includes all the mentioned topics and does a DR market analysis for both Europe and the US, being considered as an innovation and one of the purposes for answering the research question.

Databases such as IEEE Xplore, ScienceDirect, Web-of-Science, ACM, and Google Scholar will be explored for relevant articles published in the last decade. The review will focus on the evolution of DR in the US and Europe, including its adoption, implementation, and impacts on the energy markets. Additionally, case studies of successful DR programs in different regions will be examined to identify best practices and lessons learned. Finally, current developments and future perspectives of DR will be discussed, including technological advancements, regulatory frameworks, and market trends.

The paper is organized into seven sections. The introduction presents a contextualization and background about network players and the DR process, followed by the motivation and contribution of this literature review formulating the research question. Section 2 presents the literature review methodology. Next, the EU and US snapshots are presented in Section 3 and Section 4, respectively. Subsequently, the future perspectives for global DR are presented in Section 5. A comparative analysis of both these markets is conducted in Section 6. Finally, the authors summarize the findings in Section 7.

2. Literature Review Methodology

The authors conducted a systematic literature review using the PRISMA methodology [3]. The first phase involved formulating research questions to guide the review process—which was already mentioned in the previous section. The following inclusion and exclusion criteria were used during the second phase of the systematic review to obtain the research results:

• To include:
  • Describe any important DR program and related consumer concepts, such as demand side management processes and consumer flexibility for the Europe or US power and energy system.
  • Relevant documents on DR or related concepts use keywords deemed important by the authors.
  • Market analysis or case studies within the scope of Europe or US power and energy system flexibility status.
• To exclude:
  • Full paper access denied.
  • Written in a language other than English or Portuguese.

During the third phase, the authors selected online research tools and multiple databases for the review. The five selected databases were web of Science, Science Direct, Google Scholar, IEEEX, and ACM. It is important to highlight that, for a research equation formulation, the language of each tool must be considered. Science Direct, for example, does not support substituting or truncation symbols such as “?” or “*”. These symbols are useful for retrieving words with the same origin or substituting letters within a word. However, Boolean operators can be used in all other research tools to formulate equations. In addition, quotation marks indicate that the word must be present in the resulting document.

This said Figure 2 has the results obtained for each step of the systematic review information flow. It is worth mentioning that the authors considered the following keywords for the research demand response, energy market or electricity market, flexibility, and demand side management. The research finished in February 2023, so the listed references are published until now. Therefore, the studies included in this literature review have been restricted to those published within a timeframe of at most five years before the commencement of this research. This criterion ensures that only the most contemporary studies have been considered. However, older publications were referenced as support or even entity and project reports.

During the Identification stage, a substantial number of records were identified—a total of 694,693 records. Notably, this number encompasses results from five databases, multiple combinations of keywords, and a varying range of languages and geographical regions. Additionally, it is important to consider that this number may include duplicate material and papers. Therefore, dataset II removed the duplicates, non-related and other languages in the screening phase, resulting in 1496 records. The following dataset referred to the title adequacy, where 113 were chosen. Finally, regarding abstract adequacy, 50 records were selected. The following sections present the extraction, analysis, and interpretation of the information found, the discussion, and conclusions from this systematic review.

3. Europe

To speed up the process towards fully implementing the DR’s potential regarding Energy Efficiency, Directive 2012/27/EU was created to set measures to help the EU members achieve its 20% energy efficiency target by 2020 in all stages of the energy chain [26]. However, according to [27], in 2014, none of the member states had fully integrated the DR into national electricity markets due to regulation barriers: either it was not allowed, or the rules were not clearly defined, or even the existing business cases did not have the proper characteristics. Additionally, the results from this work conclude that there were 13 elements with no or residual DR participation. At that time, members such as Austria, Belgium, Finland, Greece, Ireland, and the Netherlands allowed the partial participation of consumers (principally large industrial and commercial) in balancing the market and limited wholesale. There are particular cases with actual DR in their markets, although very restricted:

• Denmark, Sweden, and the United Kingdom have the ancillary services market open to all the participants, but the wholesale and balancing markets are only open to retailers.
• France has most ancillary service markets open to all participants, but unlike the previously stated members, the wholesale and balancing markets are open to all, and the existence of aggregators is allowed.
• Germany has ancillary services, wholesale, and balancing markets open only to retailers.
• Hungary has one DR company on the wholesale and eight VPPs.
• Latvia has wholesale open.
• Poland has two programs in ancillary services open to qualified large consumers only.
• Slovenia has the ancillary service and the balancing markets open but not the wholesale. Aggregation is restricted in this country.

As can be seen, there are still many obstacles to fully implementing aggregated DR in the EU [27,28,29,30,31,32,33]. In 2016, as mentioned in [27], an open and competitive market must be required to increase the DR concept in this sector, highlighting the urgency of the full implementation of the Third Energy Package. Some suggestions for improvement were provided, such as adapting market products to DR, adapting the regulation to DR products, and enabling and empowering the demand side [28]. The authors from [27] suggested, considering the report conclusions, a template to enable aggregated DR is proposed where the market structure element and aggregation should encourage opening wholesale, balancing, and ancillary services giving the example of France as the first and only member to open wholesale market to aggregated DR [29]. Additionally, energy and availability remuneration should be paid in at least one ancillary service market, allowing investment security and promotion of participation [30]. Multiple member states already use this kind of approach. Frequent auctions, weekly and daily, update availability considering several factors that can impact their consumption profile [31]. To conquer efficiency, appropriate network fees are suggested. Consumers should not be punished by changing their load profile or participating in DR events [12].

The consumer’s capabilities should match the market’s needs, moving away from the model where the generation was the center and trying to balance the potential from both [32]. It is considered that a standardized process between Balance Responsible Parties and aggregators can be essential since creating a bridge to the energy market to these entities [33]. This way, volumes, compensation, data exchange, and governance structure are important topics for this new framework. The capabilities of the participants should be considered in the technical modalities to enable full aggregation of consumer load—abilities for participating in a market are evaluated at the aggregated pool level to be treated as a single resource and maximizing the potential, rather than for each consumer individually [34].

With this, the key factors required for the successful implementation of DR in Europe include creating an open and competitive market for DR, adapting regulations and market products to support DR, empowering the demand side, and standardizing processes between Balance Responsible Parties and aggregators [27,28,29,30,31,32,33].

With this, the work must be directed to remove this critical barrier from all Member States [27,35,36,37]. Furthermore, baseline methodology is a good option since it is appropriate and realistic and focuses on granular availability requirements. For example, taking the Austrian secondary control market, split into three periods, a consumer available during the day is paid according and does not need to be accessible at night. Besides, in the Reserve’s markets, the Short Call Duration should be 1–2 h in alignment with actual market requests [35]. Countries such as Finland, Denmark, and Sweden have already applied the no minimum capacity required for consumer participation. In the Netherlands, the real-time prices in the balancing market are communicated to consumers to allow them to react to these prices and be compensated by participating in the balance of the system. In Ireland, the Capacity market assists investment security and consumer commitment. The 2022 Europe energy crisis, triggered by Russia’s invasion of Ukraine, caused a double urgency: ending the EU dependency on Russian fossil fuels and tackling the climate crisis [36]. With this, the EU believes that resorting to market design optimization and power system flexibility can be part of the solution. So, with the 2022 REPowerEU, installing solar panels on new buildings and accelerating energy storage systems implementation becomes more than an obligation [37]. More effort is still needed to align with the NZE scenario. Although way more short and medium-term actions are meant to be applied to save the massively disrupted European and global energy markets from this war, important measures were implemented in 2022 [20], specifically in the Netherlands, where producers and consumers with more than 60 MW were required to participate in providing flexibility in heavily congested areas. Still, more efforts are needed in several aspects, as

Table 2. Overview of European implementation of DR according to the International Energy Agency (IEA) report (September 2022) [12].

TOPIC	Status
Smart Grids	More efforts
Digitalization	More efforts
Electricity Sector	More efforts needed
Buildings	Not on track
Electric Vehicles	On track
Solar PV	More efforts needed
Heat Pumps	More efforts needed

shows [12].

The importance of implementing DR measures was then addressed, highlighting the EU’s energy crisis and reducing dependency on Russian fossil fuels. Finally, baseline methodology is highlighted in [35], while [36,37] focuses on the urgency of market design optimization and power system flexibility.

Several initiatives and solutions are being created to successfully implement the DR concept in real European markets ([8,38,39,40,41,42]). Focusing on the demand side, the EcoGrid EU project creates a concept for a real-time market in the future smart distribution networks, including the high penetration of distributed generation, renewable-based technologies, and active consumer participation [38]. The idea is to reduce the need for flexibility on the production side, which can be expensive—thus balancing the power system by giving real-time and updated price signals to the active resources to provide the needed flexibility—increasing the price when there is a power deficit in the system, and vice versa. Furthermore, the InterFlex project investigates distinct flexibility market models, DSO-steered flexibility activation channels, autonomous functions and grid automation following several factors such as observed grid congestion, national regulation, and economic principles [43]. Focusing only on the use of local flexibility for the benefit of the distribution grid, it shall be recalled that flexibilities also serve other markets and purposes. From the conclusions withdrawn on their report’s Demand Response and Consumer Empowerment section, data privacy and protection rules are highlighted since access to smart metering data for input to grid constraint forecasts and flexibility procurement for grid optimization is crucial. The consumers must fill out consent forms to access any metering data, and the complexity of those is too high for residential consumers. Therefore, it is necessary to simplify without reducing the awareness of exactly how the data are employed and managed. Further studies must guarantee data protection, evaluating and adjusting the rules from data aggregation.

The DRIMPAC project aims to close the gap in the interaction between the energy market and the active participants, providing a framework that enables the end-to-end communication of essential information to provide demand flexibility and, at the same time, enhances building management intelligence mutual benefit of the prosumer and the energy system while preserving comfortable and healthy living conditions [39]. The SENDER project will develop energy service applications for DR, home automation, convenience, and security. Furthermore, with a consumer-centric approach, they engage them in a co-creation process with other players from the energy sector during the specification of pro-active DR mechanisms to supply for the consumers’ long-term motivation [44]. On the other side, the FEVER project has the objective of implementing and demonstrating solutions that control flexibility (energy storage (batteries and V2G) and DR) to give electricity network services and address problems of the distribution grid, thereby providing the EU with a secure, efficient and resilient electric grid [45].

Focusing on complex network management, the SmartNet project compared five different TSO-DSO coordination schemes—different typologies and roles for the network operators, for three national scenarios—Italy, Denmark, and Spain, where four were implemented in simulation and compared their technical and economic performance [8]. Following the new EC package Clean Energy lines, the FlexPlan project defines a new grid planning methodology where new flexibility resources and storage systems are introduced instead of grid expansion by adding new grid elements [42]. The tool integrates T&D planning, full inclusion of environmental analysis, probabilistic contingency methodologies replacing the N-1 criterion, and optimal planning decision over several decades. With this, six regional cases demonstrate the viability in real scenarios casting a view on grid planning, almost covering Europe until 2050. According to the legislative proposals made by European Commission on how the network operators must cooperate to balance the energy market and other ancillary services as well as provide congestion management, the INTERRFACE project focuses on TSO-DSO coordination enabling more efficient and effective management, increasing Demand Response and the capacity of the Renewable Generation [40]. Considering Digitalization as the key driver to achieve the goal, an interoperable pan-European grid services architecture was designed, developed, and exploited to act as the interface between the network operators and the consumers, enabling the coordinated operation of all stakeholders to use and procure common services.

Moreover, when in the presence of high shares of DG, the MERLON project developed an integrated modular local energy management framework. This project enables the comprehension and development of business models focusing on local energy communities introducing consumers to local flexibility markets [41]. Finally, the FleXunity project wants to deploy novel services for retailers and aggregators, enhanced by VPP technology empowered with artificial intelligence algorithms developed to minimize the cost of energy that can be bought in the market and optimize the use of DG from the managing entities as well as from local community’s portfolio. In addition, FleXunity promotes the active participation of community members evaluating their flexibility and energy-sharing actions [46].

While there are some overlapping themes, each reference presents unique findings and conclusions based on their specific research focus and methodology. Nevertheless, some common themes that emerge include the importance of demand-side management and flexibility in achieving a more sustainable and resilient energy system, the need for effective coordination between different stakeholders and network operators, and the role of digitalization and innovative technologies in enabling this transition. References [36,37] focus similarly on creating a framework for communication and interaction between the energy market and active participants. Refs [38,39] both focus on managing distributed energy resources. Ref [44] stands alone in its focus on deploying novel services for retailers and aggregators empowered by virtual power plant technology and AI algorithms.

4. United States

The demand response market has grown and is expected to continue and even accelerate. For years, the US has been the dominant market regarding DR, mainly due to the application of new policies that allow players of all sizes to participate in wholesale electricity markets, covering 60% of the US power supply [47]. In this case, residential consumer participation has the highest amount—around 90%, but the remaining sectors provide the major market share regarding flexibility, incentives, and savings. In 2016, California was the most active state in the US regarding DR markets—20% of the total Demand Response customers in the US and contributing 20% of the total peak demand savings. However, throughout the years, DR participation in the wholesale market decreased by 1.383 MW from 2019 to 2020 [48].

Parameters and specifications must be considered to create a successful DR program. Considering an applied example of DR programs on the wholesale markets administered by ISO and RTO in North America. The Federal Energy Regulation Commission (FERC) updated the North American Energy Standards Board business practice standards for measuring and verifying demand response and energy efficiency on 21 February 2013 [3]. Common definitions and processes were provided regarding DR products/services. Additionally, the applicability of performance evaluation requires ISO and RTO to file revised governing documents regarding the assessment methods to be used on DR products/services. Four categories are considered in the applicability of measurement and verification standards:

• Energy Service—Demand resources deliver a quantity of electricity, measured in MWh;
• Capacity Service—Demand resources are required as a means of managing demand over a defined period, measured in MW;
• Reserve Service—Demand resources are obligated to be available to produce reduction upon deployment period;
• Regulation Service—Demand Resource fluctuates load in response to real-time signals from the System Operator. These resources are subject to dispatch continuously during the commitment period.

Providing an overview of selected appliable features,

Table 3. Overview of Product/Service Features from DR programs applied in North America—Capacity and Energy.

	Min. Eligible Resource Size	Min. Reduction Amount	Agg. Allowed	Response Required	Trigger Logic	Total DR Contribution Limit (%)	Min. Sustained Response Period	Max. Sustained Response Period	Max. Deployments per Availability Window	Obligation Period	Availability Window	Demand Resource Availability Measurement
Capacity	100 kW – 1 MW	1 kW – 500 kW	Yes (12) No (1)	Mandatory	Operational (8) Other (5)	-	- (8) 5 min–4 h (5)	- (5) 3 h–24 h (8)	- (8) 1–8 (5)	All year (1) ERS Periods Awarded (4) Seasonal (6) Schedule (1) - (1)	All hours (1) ERS Periods Awarded (4) Seasonal (6) Schedule (1) - (1)	Annual test (1) Calculated after the commitment period (4) Daily update (1) Telemetry (2) - (3)
Energy	100 kW – 1 MW	10 kW – 100 kW	Yes (13) No (2)	Mandatory (8) Voluntary (6)	Operational (6) Other (9)	-	5 min–4 h (6) - (7) EDR offer (1) Offer (1)	4 h (2) - (6) Based on offer (1) Based on capacity (1) Dep. window (3) EDR Offer (1) Schedule (1)	- (10) 1 (1) 12 (1) Based on offer (1) Based on component (1) EDR offer (1)	- (3) All hours (1) Based on component (1) EDR offer (1) Schedule (6) Seasonal (3)	- (3) All hours (1) Based on capacity (1) Dep. window (1) EDR offer (1) Schedule (7) Seasonal (1)	- (4) As bid (1) Annual test (1) Daily update (1) ICCP (1) Offers (4) Telemetry (3)

and

Table 4. Overview of Product/Service Features from DR programs applied in North America—Regulation and Reserve.

	Min. Eligible Resource Size	Min. Reduction Amount	Agg. Allowed	Response Required	Trigger Logic	Total DR Contribution Limit (%)	Min. Sustained Response Period	Max. Sustained Response Period	Max. Deployments per Availability Window	Obligation Period	Availability Window	Demand Resource Availability Measurement
Regulation	100 kW – 1 MW	100 kW – 1 MW	Yes (3) No (2)	Mandatory	Automatic (1) Operational (2) Other (2)	- (4) 25% (1)	- (3) 1 h (2)	- (3) Dep. window (1) Schedule (1)	-	All hours (1) Schedule (4)	Schedule	ICCP (1) Telemetry (3) Offers (1)
Reserve	100 kW – 10 MW	100 kW – 10 MV	Yes (9) No (5)	Mandatory	Operational (6) Other (8)	- (9) 50% (1) 33% (1) 25% (1) 40% of spin requirement for DDr (2)	1 h (5) - (9)	- (9) Dep. window (2) Schedule (2) 30 min (1)	-	Contract (2) Schedule (7) - (2) All hours (2) Between arming and disarming (1)	- (3) Contract (2) Schedule (9)	Actual, offered and armed volumes reported (1) ICCP (1) Offers (4) Telemetry (8)

summarize, according to the service type, the active DR programs resorting to information provided in “North American Wholesale Electricity Demand Response Program Comparison,” updated on November 2018 [49]. It must be highlighted that Table 3, Table 4,

Table 5. Overview of Event Timing from DR programs applied in North America [48].

	Advance Notification(s)	Ramp Period	Sustained Response Period
Capacity	None (7) Day-ahead (3) 5 min–2 h (2) Defined in Market Rules (1)	5 min–2 h (8) Effectively Instantaneous (2) Resource-Specific (1) Included in energy market offer (1) N/A (1)	As Dispatched/Recalled (7) 5 min–8 h (6)
Energy	None (2) Day ahead (9) 5 min–2 h (4)	5 min–2 h (9) Based on Resource Parameters (1) Startup time and ramp rate included in energy market offer (1) Resource Specific (2)	As Scheduled/Dispatched (9) 5 min–4 h (6)
Regulation	None (2) Day ahead (2) 5 min (1)	Effectively Immediate (4) 4 s (1)	As Scheduled/Dispatched (4) 10 s to 60 min (1)
Reserve	None (5) Day-ahead (4) 5 min–2 h (4) real-time (1)	Ramp rate include in the energy offer (1) 0.2 s–30 min (13)	As directed (1) As dispatched (8) 5 min–1 h (5)

and

Table 6. Performance Evaluation Methodologies.

Performance Evaluation Type	Service Type
	Energy	Capacity	Reserves	Regulation
Baseline Type-I	✓	✓	✓
Baseline Type-II	✓	✓	✓	✓
Maximum Base Load	✓	✓	✓
Meter Before/Meter After	✓	✓	✓
Metering Generator	✓	✓	✓	✓

have represented between curly brackets the number of DR programs that fulfill the specific characteristic.

By analyzing Table 3 and Table 4, the minimum eligible resource size considered varies between 10 kW and 100 MW, but the minimum reduction amount lies between 1 kW and 10 MW. The majority allows aggregation. Only in the Energy service, the response required can be voluntary. There are several trigger logics, but the main ones are operational procedures or automatic responses. The total contribution limit was only applied to Regulation and Reserve services fluctuating between 25% and 50%. Many DR programs do not restrict the Sustained Response Period with a minimum and maximum, but one example goes between 3 h and 24 h. Energy and Capacity services have a maximum number of deployments per availability window. A low percentage of DR programs do not contemplate an availability window.

Regarding the Demand Resource Availability Measurement, Telemetry has the highest percentage among all the programs. The DR timeline is also important; as presented in Figure 1, each period must have its own rules. Contemplating the same logic as Table 3 and Table 4, the DR programs per each service type are analyzed regarding event timing: advance notification, ramp period, and sustained response period.

Regarding further notice, none is the most frequent, followed by day-ahead until 5 min. The ramp period is immediate to 2 h, mostly for DR programs. Additionally, it can be included in the energy offer or resource specific. For the Sustained Response Period, it normally happens as scheduled/dispatched. Table 3, Table 4, Table 5,

Table 7. Performance evaluation—Baseline Type I and Baseline Type II.

	Baseline Information (Baseline Window and Calculation Type)	Real-Time Telemetry	After-the-Fact Metering	Performance Window	Measurement Type
Baseline Type -I	Model built using historical meter data (12+ months of historical data)	Yes (5) No (12)	Yes (15) Optional (2)	Sustained Response Period (11) Event-dependent, as specified in Notification instructions (2) Sustained Response period or optionally Deployment Period (Participant Selection (2) 5 min and hourly (2)	15-min Interval Data Recorder (6) 5-min interval load (2) Hourly metered load (4) Actual vs Setpoint (2) Customer/Resource specific (1)
	Average - 10 most recent like days (exclude highest and lowest of the 10 most recent like days) - 20 most recent like days - 45 days (10 most recent non-event, like days)
	Best matching event days and day preceding event day from prior 12 months
	Actual metered load to control group sample average
	Compare metered interval load during the deployment to the load of the 15-min prior to issuance of deployment
	5-min load data of qualifying days: 90 % of the prior qualifying baseline + 10% of the previous qualifying day
	For each baseline day type, calculate the rolling average of a 5-min load from the most recent days on which the resource was not dispatched: - 10 days for non-holiday weekdays - 5 days for Saturdays - 5 days for Sundays and holidays
	Customer/Resource Specific
	Weekdays: Hourly simple average of the 5 highest total event period load days in CBL Window (10 previous weekdays within the last 30 days, subject to exclusion rules --> exclude day preceding event/holiday and curtailment events) Weekends: Hourly simple average of the 2 highest total event period load days in CBL Window (previous 3 weekends—same day type (no exclusions))
	Hourly average based on high 4 of 5 days weekdays and high 2 of 3 for Saturday or Sun/Holidays (45 days)
	Alternative calculations as long as it will significantly improve accuracy compared to standard method
	Avg. hourly integrated DR load for the same hours in the last 30 calendar days when the resource was not dispatched, adjusted when events accur
Baseline Type II	Baseline window/Calculation Type define for other resources, as approved on a case by case basis	Yes (1) No (5)	Yes (5) Optional (1)	Sustained Response Period (5) Event-dependent (1)	Statistical equivalent of 5 min or hourly metered load (5) Customer/Resource Specific (1)
	Customer/Resource Specific
	Simple Average (45 Days) --> (exclude the 10 most recent non-event, like days)
	Approved on case by case basis or may use published deemed savings study

and

Table 8. Performance evaluation—Maximum Base Load, Meter before/after, and Metering.

	Baseline Information (Baseline Window and Calculation Type)	Real-Time Telemetry	After-the-Fact Metering	Performance Window	Measurement Type
Maximum Base Load	Average Coincident Load (ACL): Avg. of the top 20 out of the top 40 coincident peak hours from the Prior Equivalent Capability Period. Coincident peak hours exclude DR events. Capacity only.	Yes (2) No (4)	Yes (5) No (1)	Sustained Response Period (5) Event-dependent (1)	SCADA or Meter 15-min data vs. max. baseload 5-min interval load Customer/Resource Specific Hourly meter data Av. performance window
Meter Before/Meter after	Single Reading 0.2 s after frequency drops to 5.9 Hz for LSSi and 10 min ater Directive for SUP Meter read before deployment Unit special processing—sustained response period (1 h) Unit Special processing—deployment (1 min) Unit Special processing—2 s Scan Rate following signal	Yes (7) No (4)	Yes (9) Yes, interval meter data is collected (1) No (1)	Sustained Response Period (10) Any hours obligated in Reg.-Up or Reg.-Down (1)	5-min interval load (1) Actual vs. Setpoint (2) Telemetry (2) Customer/Resource Specific (1) Host load forecast—integrated 1-min meter data (1) Instant. load (1) Avg. performance window (3)
	Meter read before deployment and sustained through the deployment (For Spin/Non-Spin no-pay calculation)
	Actual telemetered load vs. 5-min average telemetered load prior to event
	Actual telemetered load vs. dispatched set point (45 s)
	Single reading (Meter read before deployment)
	1-min data (before deployment + Host Load Zone Forecast)
Metering Generation Output	Difference between actual metered output during event and generator’s typical use hour (calculated via 10-in-10 baseline) (Typical use baseline—45 Days)	No	Yes, interval meter data is collected	Sustained Response Period	5-Min load

refer to a summary of the outcomes of [49].

To assess the performance of each Demand resource, some rules must be followed. The process refers to the methodology to estimate the DR product/service’s reduction value. The following five methods are the main choices unless others are specified by [50]:

• Baseline Type I (interval meter)—based on Demand Resource’s historical interval meter data (may include other variables such as weather and calendar data);
• Baseline Type II (non-interval meter)—based on statistical sampling to estimate the usage of a Demand Resource, considering that an interval metering is not available on the entire aggregated population;
• Maximum Base Load—based on Demand Resource’s ability to maintain usage at or below a specified level during a DR event;
• Meter Before/Meter after—based on a comparison between electricity demand over a specified period preceding deployment and similar readings during the sustained response;
• Metering Generator Output—considers the Demand Reduction Value as the output of a generator located behind the Demand Resource’s revenue meter.

Considering the four service types, Table 6 indicates which performance evaluation methodology can be employed for each one throughout check marks, and the information is completed with Table 7.

Centering on the DR programs applied in the North America Wholesale market, Table 7 and Table 8 summarize some methods used for performance evaluation based on Baseline Information, use of real-time telemetry or/and After-the-Fact Metering, performance window, and measurement type. The baseline is a method to foresee the likely load curve from a Consumer or Demand Resource in the lack of a DR event and can be used in all service types. An adjustment can be applied to reflect the real circumstances immediately before or during a DR event regarding but not limited to weather conditions, near event facility load, or other parameters. As mentioned, interval metering (Type I) or statistical sampling (Type II) can be calculated. The calculation type for developing the baseline should be specified by Governing Documents, for both types of maximum, average, and regression can be applied, not being exclusive.

The energy transition will continue to grow in 2023, but significant challenges must be overcome. Starting with the Inflation Reduction Act (IRA), which includes numerous tax credits and other incentives, the clean energy sector is anticipated to receive a significant boost [51]. Deployments of renewable energy and energy storage are also being pushed forward by increasingly ambitious decarbonization targets set by governments and businesses. The FERC, the Department of Energy, grid operators, and others also deal with many other problems. For instance, developing a solution to build more transmission and make it easier to connect new resources to the grid, interregional transmission planning, and the threats that extreme weather poses to grid reliability are among the other issues it may address [48].

Still, FERC will likely continue focusing on grid reliability and resilience. In the meantime, sales of electric vehicles are growing quickly. However, the IRA’s tax credits and the construction of a nationwide charging station network raise important questions for the industry.

Cybersecurity also becomes an important topic to discuss since new regulations might be applied amid the rise in physical attacks. In 2022, cybersecurity and the potential threat that hackers may pose to the electric grid’s reliability were discussed during the war in Ukraine and the ongoing rise in ransomware.

However, in December, the industry’s attention was drawn to two physical attacks that caused thousands of outages on both sides of the United States. As a result, the Department of Energy is funding “next generation” cybersecurity research, development, and demonstration projects, and FERC is considering developing new cybersecurity rules for distributed energy resources on the bulk electric system [52].

After experiencing robust growth in 2022, energy storage is anticipated to expand significantly in 2023 by almost any measure. The need for storage to store energy produced by intermittent resources such as wind and solar is rising as governments at all levels and businesses broaden their goals for reducing carbon emissions. According to the US Energy Information Administration, developers and power plant owners intend to nearly quadruple the utility-scale battery storage capacity in the US over the next three years, reaching 30 GW by the end of 2025 [53].

Large-scale battery storage projects in the US are also expected to benefit from the IRA. Standalone storage systems will be eligible for a 30% investment tax credit and up to 70% with additional incentives. It is not necessary for batteries to be paired directly onsite with solar PV generation to be eligible for an approximately 30% cost reduction thanks to the investment tax credit subsidy for standalone energy storage. In the past, energy storage projects could only claim the tax credit if they were installed alongside a brand-new solar generation facility and were charged at least 80% by the solar facility [54].

Focusing on EVs, experts in the field predict that sales of electric vehicles could double again in 2023 [55]. They say that the next year will be crucial to continuing the trend of transportation electrification as billions in federal incentives begin to hit the market, even though that growth trajectory is not sustainable over the long term. In addition to other investments, the IRA extended federal tax credits for vehicle purchases, and the bipartisan infrastructure law of 2021 provided $7.5 billion for a national network of 500,000 electric vehicle chargers [56]. Analysts believe the speed with which the United States can achieve its objectives will depend on how these incentives are implemented.

5. Future Perspectives

The rising proportion of renewable energy sources in the mix will significantly affect the design of electricity markets. The wholesale price of electricity is often zero or even negative when the proportions of solar, wind, other variable renewables, and nuclear power reach high levels. This is because the amount of electricity that is available at no marginal cost is frequently higher than the amount of electricity that is needed. Net Zero Emissions (NZE) by 2050 Scenario is the world’s first comprehensive study of how to transition to a net zero energy system by 2050 [20]. The Intergovernmental Organization for Energy Efficiency (IEA) strives to provide everyone with affordable, clean, and reliable energy. It has 31 member countries, including the United States, Japan, Germany, the United Kingdom, and France. This entity aims to foster economic expansion, environmental sustainability, and energy security, promoting the transition to a more sustainable energy future by providing analysis, data, and recommendations on energy policy and energy markets.

In one of the more recent analyses from the IEA, despite the active use of demand response, approximately 7% of wind and solar output in the NZE would be above and beyond what can be integrated by 2050. The share of zero-price hours in the year would increase to approximately 30% in major markets from close to zero today [12]. It would, therefore, be highly desirable to make significant changes to the design of electricity markets to provide signals for investment, including investment in sources of flexibility such as battery storage and dispatchable power plants, to achieve the NZE’s goal of increasing the proportion of renewables in the mix that generates electricity. Yet, there is a long road to implementing the DR in the real world fully. For DR, a positive trend continues rising in the regulation and implementation in several countries since 2020, namely the development of existing approaches and a tentative for enabling the participation of smaller resources in market transactions. Still, the pace of policy implementation and technology deployment should increase to achieve the goal. By 2030, this scenario has a DR milestone of 500 GW brought onto the market—ten times the deployment levels in 2020 [12]. According to the developed scenario, DR and energy storage systems combined should meet a quarter of the flexibility needed globally by 2030—more than half of the flexibility needed by 2050.

Regarding coal and oil, advanced economies should disappear to be substituted by a mix of hydrogen-based, natural gas, nuclear, hydro, and other renewables [12]. Furthermore, it is expected that the consumer side becomes more aware of the market’s transactions—not only by shifting consumption to periods when renewable energy is abundant (through a more conventional source of DR) but also by supplementing new resources such as the smart charging of EVs (unlocking new valuable ways of DR). Actually, EVs provide a crucial portion of the overall system flexibility. However, although already applied in the power and energy system, the evolution has been slow due to regulatory and institutional difficulties that are expected to be overcome in the NZE.

Nonetheless, experiments are being conducted around the world to achieve NZE goals. Countries such as France, Italy, the Netherlands, and the United States (US) have increased vehicle-to-Grid (V2G) charging experimentation. In 2022, the United Kingdom (UK) started a program to test the interoperable DR using smart meters and energy management systems. Regarding Virtual Power Plants (VPP), several projects are being launched in Australia, with initiatives and pilots multiplying in Western Australia, Victoria, and New South Wales. However, in 2021 only 31 MW were enrolled in Australian Energy Market Operator VPP [57]. Still, companies are preparing for the future and betting on this approach. For instance, Tesla has been expanding its VPP in South Australia and the US (California and Texas) [58], and Stem Inc. is developing the first one in Latin America, located in Chile [59].

Since 2020, European markets have increased their capacity for demand response, with some nations launching their first auctions or diversifying their portfolio of demand-side resources [20]. For the second year in a row, the French market for demand-side flexibility is expanding rapidly. In 2022, selected bids will total 2.4 GW, an increase of 1 GW from the levels in 2021. In its one-year-ahead auction, the United Kingdom secured 528 MW of demand-side resource capacity in 2022, more than doubling capacity in 2021. With 1 GW of demand-side resource capacity awarded, the four-year-ahead auction remained at 2021 levels. Korea’s demand response markets had approximately 4.55 GW of registered capacity as of January 2022. This capacity is especially important for reducing peak demand during the summer and winter months. In December 2021 alone, the economic demand response program prevented the consumption of 29 GWh. In its first capacity auction four years in advance, Belgium selected 4.45 GW of total capacity in October 2021; However, demand and storage account for only 8% of the selected offers.

In 2022, Australia introduced the Five-Minute Settlement in the national electricity market, which provides more effective price signals and encourages battery investment. The new wholesale DR mechanism allows consumers to sell DR. In anticipation of decommissioning nuclear power plants, Belgium launched the first auction of a four-year ahead capacity remuneration mechanism to bolster supply security [20]. With the 2019 Energy Pause program (residential demand response for small consumers below 70 kW), the 2020 Fast DR program to improve frequency stability, and the 2021 Plus DR Program to minimize renewable energy curtailment, Korea expanded its demand response programs. As a step toward making demand response programs easier, Colombia approved the implementation of advanced metering infrastructure and a regulatory roadmap. Additionally, India approved guidelines to make it easier for energy storage to participate in the grid’s provision of flexibility and services.

Regarding the US, it is believed that power systems will face three primary grid reliability challenges in 2023. First, unpredicted extreme weather, particularly heat waves, can result in high electric loads, wildfires, and storms that can harm generators and transmission lines. The development of resources to replace the retiring generation can also be slowed down by ongoing issues with the supply chain. In different ways, utilities and regulators are tackling these issues. Several utilities are analyzing how high their electric loads could be in response to potential heat wave scenarios and updating their load forecasts [60].

Through IRA, the renewable energy sector received a significant boost. This means 2023 is considered a year of growth and continued business challenges as companies take advantage of billions of dollars in tax credits. Due to ongoing disruptions in the global supply chain and the Uyghur Forced Labor Prevention Act, which caused some solar module imports to be held at the border, the solar industry experienced declining growth throughout 2022. Growth was also affected by uncertainty about future imports due to a Commerce Department investigation into tariff evasion.

6. Comparative Analysis

As mentioned earlier, the implementation of DR in Europe faces regulatory barriers in some member states, which limits its full integration into national electricity markets. It is then suggested to enable aggregated DR, from the authors’ perspective, the development of a business model that includes opening wholesale, balancing, and ancillary services markets, allowing investment security and promotion of participation, and adapting market products and regulations to DR. Standardizing processes between BRP and aggregators can be essential, along with evaluating participants’ technical capabilities at the aggregated pool level. Baseline methodology and real-time prices in the balancing market are suggested, along with more effort needed to align with the Net Zero Emissions scenario. Regarding regulatory barriers, such as member states’ individual energy policies and lack of cooperation, can delay the full integration of DR in the EU. Standardization and alignment with the Net Zero Emissions scenario are seen as ways to address these issues. Additionally, the 2022 Europe energy crisis has triggered a double urgency: ending the EU dependency on Russian fossil fuels and tackling the climate crisis—market design optimization and power system flexibility can be part of the solution.

In contrast, the DR market in the US has been dominated by policies that allow players of all sizes to participate in wholesale electricity markets. California was the most active state in the US regarding DR markets in 2016. Regarding energy storage and electric vehicles, the growth trajectory is not sustainable over the long term. Analysts believe the speed with which the United States can achieve its objectives will depend on how these incentives are implemented.

By comparing both perspectives, the authors suggest that Europe can learn from the US policies that have allowed players of all sizes to participate in wholesale electricity markets, which has contributed to the growth of the DR market in the US. Europe can also learn from the US in terms of incentivizing large-scale battery storage projects. However, the authors alert against directly applying US policies to the EU, as the EU has different energy policies and regulations. It is also recommended that Europe should not follow the US in relying solely on federal incentives to drive growth in the DR market and transportation electrification. Instead, the EU should focus on standardizing processes and regulations, promoting investment security, and aligning with the Net Zero Emissions scenario.

In conclusion, while Europe faces regulatory barriers to the integration of DR, the region has made significant progress in recent years, with the EU targeting a 55% reduction in greenhouse gas emissions by 2030. The EU can learn from the US approach to DR market development, investment in large-scale battery storage projects and electric vehicle infrastructure, and federal incentives for the adoption of renewable energy technologies. However, Europe should also be cautious about replicating aspects of the US approach that may not be suitable for the region’s unique regulatory and market structures. By optimizing market design and increasing power system flexibility, both the US and Europe can contribute to achieving their energy and climate goals.

7. Conclusions

A key strategy for balancing supply and demand in electricity grids and enabling more efficient and sustainable use of energy resources is most certainly DR. Although the study was made from a separate perspective, the US and EU both recognize the significance of DR as crucial for balancing supply and demand in electricity grids and enabling more efficient and sustainable use of energy resources. The implementation of DR is somewhat similar. However, DR programs and technologies have progressed more rapidly in some parts of the United States than in other parts of Europe. In fact, the regulatory and policy environment, which has been more favorable for DR in some regions of the United States, contributes to this distinction. Furthermore, the availability of enabling technologies, such as smart meters and smart appliances, which have been more widely implemented in the US, is another crucial factor contributing to the disparity in DR implementation. The US has made it easier for consumers to participate in DR programs and obtain real-time information about how much electricity they use and how much it costs.

In recent years, both Europe and the United States have made efforts to combat climate change. The European Union has taken a leading role in this effort, setting ambitious targets to reduce greenhouse gas emissions and promoting renewable energy sources. The EU has committed to reducing greenhouse gas emissions by at least 55% by 2030 compared to 1990 levels, and to reach net zero emissions by 2050. To achieve these goals, the EU has implemented a range of policies and initiatives, including the EU Emissions Trading System, the Renewable Energy Directive, and the Energy Efficiency Directive.

Similarly, the United States has also taken steps to address climate change, and proposed a range of measures to accelerate the transition to clean energy, including investing in renewable energy and electric vehicles, improving energy efficiency, and phasing out fossil fuel subsidies.

Despite these differences, efforts are being made for DR programs and technologies related to being developed and implemented in both the US and Europe. From the discussion, it is possible to conclude that our ongoing efforts to harmonize policies and regulations, encourage innovation, and invest in creating new demand response technologies are more effective. Overall, while there is still much work to be conducted, Europe and the United States have shown a commitment to tackling climate change and are taking concrete steps to reduce greenhouse gas emissions and transition to a more sustainable future. While the US has a larger experience and actual implementations, it is not possible to easily find a clear reason why Europe is seeking DR with lower levels and late in time.