The Power Regulation Characteristics, Key Challenges, and Solution Pathways of Typical Flexible Resources in Regional Energy Systems

Abstract

The low-carbon transition of the global energy system is an urgent necessity to address climate change and meet growing energy demand. As a major source of energy consumption and emissions, buildings play a key role in this transition. This study systematically analyzes the flexible resources of building energy systems and vehicle-to-grid (V2G) interaction technologies, and mainly focuses on the regulation characteristics and coordination mechanisms of distributed energy supply (renewable energy and multi-energy cogeneration), energy storage (electric/thermal/cooling), and flexible loads (air conditioning and electric vehicles) within regional energy systems. The study reveals that distributed renewable energy and multi-energy cogeneration technologies form an integrated architecture through a complementary “output fluctuation mitigation–cascade energy supply” mechanism, enabling the coordinated optimization of building energy efficiency and grid regulation. Electricity and thermal energy storage serve as dual pillars of flexibility along the “fast response–economic storage” dimension. Air conditioning loads and electric vehicles (EVs) complement each other via thermodynamic regulation and Vehicle-to-Everything (V2X) technologies, constructing a dual-dimensional regulation mode in terms of both power and time. Ultimately, a dynamic balance system integrating sources, loads, and storage is established, driven by the spatiotemporal complementarity of multi-energy flows. This paper proposes an innovative framework that optimizes energy consumption and enhances grid stability by coordinating distributed renewable energy, energy storage, and flexible loads across multiple time scales. This approach offers a new perspective for achieving sustainable and flexible building energy systems. In addition, this paper explores the application of demand response policies in building energy systems, analyzing the role of policy incentives and market mechanisms in promoting building energy flexibility.

1. Introduction

1.1. Background

With the continuous rise in global energy demand and the escalating climate crisis posing serious challenges to human survival, renewable energy has gradually become a central focus of the energy transition. As a key component of a clean energy system, renewable energy offers a critical pathway to address climate issues, which is exemplified by China’s ongoing efforts to build a new power system dominated by renewables [1]. However, due to the influence of stochastic variables such as weather conditions and seasonal changes, renewable energy output is characterized by high intermittency and significant volatility. In scenarios with a high penetration of renewable energy, emerging power systems urgently need to overcome technical bottlenecks in maintaining a real-time supply–demand balance [2]. On the generation side, traditional power sources suffer from slow regulation speeds, and the inherent variability of renewable energy—combined with insufficient energy storage capacity—limits flexibility in meeting real-time demand. In contrast, the demand side offers abundant flexible energy resources. Leveraging distributed flexible resources such as building energy systems to enhance grid stability is considered a cost-effective solution with relatively low marginal costs [3,4].

The concept of building energy flexibility has multiple sources. The International Energy Agency (IEA) defines it as “the ability of a building to manage its demand and generation according to local climatic conditions, user needs, and grid requirements” [5]. In China, the “Assessment standard for Photovoltaics, Energy storage, Direct current and Flexibility (PEDF) system in buildings” (T/CABEE 055-2023) [6] defines it as the capability of buildings and their occupants to actively regulate power consumption through electrical equipment, electrochemical energy storage, thermal (or cooling) storage, the thermal inertia of building envelopes, or adjustments in user behavior. Despite differences in phrasing, these definitions converge on a common focus: reshaping electricity demand curves through two fundamental pathways—user demand elasticity and multi-energy substitution—to reduce the interaction power with the grid [6]. User demand elasticity primarily refers to modifying energy consumption patterns within a flexible demand range to enable load modulation, such as adjusting indoor air conditioning loads based on thermal comfort tolerance [7], or leveraging charging time flexibility to modulate EVs loads [8]. Multi-energy substitution involves the equivalent transfer between different energy forms, such as using distributed renewable generation or distributed storage—including thermal storage [9], cold storage [10], and battery storage [11]—during peak electricity demand periods to reduce dependence on grid-supplied electricity. Together, these two mechanisms form the core realization pathway of building energy flexibility, offering critical technical support for supply–demand coordination in next-generation power systems.

Flexible resources in buildings can be classified across multiple dimensions, including regulation characteristics, application scenarios, response time, and energy types. Based on regulation characteristics, they can be categorized into interruptible, shiftable, and adjustable loads. Interruptible loads refer to those that can be completely shut off or significantly reduced during specific periods, such as non-continuous industrial equipment [12,13], certain air conditioning systems [14,15], and non-critical lighting [16,17]. Shiftable loads are those whose electricity usage time can be flexibly adjusted without changing total energy consumption, including appliances such as washing machines, dishwashers, and EVs chargers [18,19,20]. Adjustable loads refer to loads whose power consumption can be dynamically modulated, including variable-speed air conditioners, water pumps, and battery energy storage systems (BESS). According to application scenarios, flexible resources can be classified into industrial, commercial, and residential categories. Industrial buildings are characterized by high power demand, strong regulation potential, low climate sensitivity, and high automation, but must maintain process continuity [21]. Typical flexible resources include schedulable or interruptible loads within energy-intensive processes. Examples range from batch-processing equipment, such as electric arc furnaces in steelmaking and kilns or grinders in cement production, to ancillary systems like compressed air and large-scale ventilation in mining. Steel plants adjusting furnace operation schedules for peak shaving or paper mills curtailing load by shutting down grinders are classic instances of leveraging this potential [22,23,24]. Commercial building flexibility is closely tied to business hours and requires consideration of comfort levels; typical systems include central air conditioning, lighting, and elevators, with applications such as shopping malls adjusting air conditioning setpoints during peak periods [25], lighting systems employing hybrid daylight–artificial light strategies [17], and cooling systems participating in demand response via optimized compressor operation [15]. Residential building flexibility is characterized by high randomness and decentralization, with small individual loads but large aggregate potential that relies heavily on user behavior analytics; typical devices include water heaters, residential storage systems, and smart plugs, with applications such as smart home systems that optimize load schedules [26], baseload modulation by air conditioners and water heaters [27], and load shifting by washing machines and similar appliances [18]. In terms of response time, flexible resources can be divided into second-level, minute-level, and hour-level resources. Second-level resources, such as storage systems and uninterruptible power supplies (UPSs), can respond rapidly to frequency fluctuations and participate in primary frequency regulation [28]; minute-level resources, such as air conditioning and water pumps, respond within several minutes to half an hour and are commonly used in demand response markets [29]; hour-level resources, such as EV charging and industrial processes, operate on timescales of hours to days and are suited for load shifting under time-of-use electricity pricing [30]. Based on energy types, flexible resources can be classified as electric, thermal, or hybrid. Electric resources typically include air conditioning, lighting, and electric motors; thermal resources include thermal storage tanks, heat pumps, and district heating systems; hybrid resources include combined heat and power (CHP) systems and integrated energy systems (IES).

A systematic classification and analysis of the flexibility characteristics of building-based flexible resources form the foundation for the deep participation of demand-side resources in coordinated grid regulation. However, practical implementation faces compounded challenges arising from the heterogeneity of building types, the diversity of spatial scales, and the integration of emerging resources. Among these, building-type heterogeneity is a core influencing factor that directly determines the regulation characteristics and technical pathways of flexible resources in different building scenarios [5]. Public buildings are primarily characterized by adjustable air conditioning and lighting loads, which respond quickly but are constrained by occupant comfort requirements [17,31]; industrial buildings can achieve substantial flexibility through interruptible processes and large-scale equipment. However, harnessing this potential is complex, as it requires dynamic coordination with production schedules. Furthermore, the flexibility characteristics of industrial loads are highly heterogeneous, being deeply dependent on specific processes (e.g., steelmaking, chemical production, cement grinding) [32]. A detailed comparative analysis is therefore beyond the scope of this review, which focuses on buildings and EVs; residential buildings exhibit highly dispersed resources and significant behavioral uncertainty, with effective participation relying on systematic design of intelligent control technologies and incentive mechanisms [33]. Spatial-scale differences are reflected in the decentralized nature of single-building regulation capabilities versus the collaborative advantages of resource aggregation, where the latter can enhance overall load response robustness through spatiotemporal complementarity mechanisms. More critically, the large-scale integration of EVs as mobile flexible resources is profoundly reshaping the architecture of building energy systems. Their dynamic charging and discharging characteristics and bidirectional energy interaction capabilities introduce new uncertainties while simultaneously expanding the potential for multi-energy synergy [34,35]. Therefore, it is essential to systematically map building energy flexibility resources and their response characteristics from multiple dimensions—such as functional attributes, spatial hierarchies, and temporal scales—to enable efficient integration of demand-side resources and support coordinated grid regulation.

1.2. Previous Reviews

Existing reviews have systematically examined building flexibility resources with a focus on specific building functional types, individual flexible resources, and different spatial scales.

In terms of building functional types, existing studies have predominantly focused on residential buildings. For example, Luo, Z. et al. [36] established a quantitative framework for energy flexibility in residential buildings through systematic conceptual clarification, delineating the theoretical boundaries among demand flexibility, operational flexibility, and energy flexibility, and constructing a methodological framework covering definitions, load classification, and quantitative evaluation. Rajendhar, P. et al. [37] examined the engineering feasibility of Home Energy Management Systems (HEMSs) from a technological implementation perspective, addressing key components such as intelligent control algorithms, hardware architecture design, communication protocol adaptation, and demand response optimization, thereby providing an integrated solution for managing energy flexibility in residential contexts. Lind, J. et al. [38] systematically reviewed the application of building thermal mass (BTM) in residential buildings, highlighting its significant thermal storage potential and load-shifting capability while also analyzing the coupling effects of heating system configurations and environmental parameters on BTM capacity, offering theoretical support for thermal–electric interaction in buildings. Li, H. et al. [39] focused on the characteristics and quantification methods of energy flexibility in residential buildings, developing a comprehensive framework including load classification, flexibility metrics, and application scenarios. Their analysis of methodological boundaries in current quantification approaches revealed research gaps in multi-timescale coordination and user behavior modeling, thus providing a theoretical foundation for the large-scale development of flexible residential resources. Pallonetto, F. et al. [40] conducted a systematic review of residential DR implementation pathways, proposing a flexibility assessment method based on linear optimization models. They emphasized the impact mechanisms of market incentive design and user acceptance on DR effectiveness and validated the implementation performance of DR under different market models through case studies, offering methodological insights for the practical implementation of residential DR programs.

At the flexible resource level, the existing literature typically focuses on two representative categories: EVs and temperature-controlled loads. For instance, Kakkar, R. et al. [41] systematically reviewed the technical landscape of EV participation in demand response, covering key elements such as classification models, load forecasting algorithms, and optimization control strategies. Di Silvestre, M.L. et al. [42] investigated the flexibility potential of electric water heaters under the U.S. market context, providing a comprehensive analysis of their operational characteristics and control system architecture, and quantitatively demonstrating their significant capacity for system-level flexibility regulation. Nema, S. et al. [43] conducted a systematic review of control strategies for temperature-controlled loads (TCLs) participating in fast frequency response (FFR), establishing a general TCL modeling methodology, clarifying the core terminology in the TCL control domain, and exploring standard aggregation control strategies and methods. Wang, H. et al. [44] focused on centralized Heating, Ventilation and Air Conditioning (HVAC) systems in non-residential buildings, systematically summarizing their participation in ancillary grid services such as frequency regulation and spinning reserve, while also analyzing technical pathways and quantitative assessment methods. The study further examined challenges related to system compatibility and control precision in real-world applications, providing a methodological reference framework for the large-scale aggregation and control of non-residential HVAC systems.

At the spatial scale, existing studies have primarily focused on individual buildings. For example, Bai, Y. et al. [45] summarized the operational characteristics of flexible resources on both the supply and demand sides within single buildings and proposed response strategies for the coordinated utilization of various flexibility resources. They suggested that the deployment of flexible resources should consider building typology and resource characteristics to maximize demand response potential. Building upon this, Chen et al. [46] developed an integrated analytical framework for grid-interactive buildings, which not only introduced a strategy matrix for enhancing energy flexibility but also proposed a universal methodology for evaluating demand-side flexibility. Sharda, S. et al. [26] addressed the practical challenges of coordinating multiple household appliances for demand-side management (DSM) within residential buildings through HEMSs. They systematically analyzed the multi-dimensional technical architecture of DSM optimization, covering key components such as load characteristic modeling, the integration of distributed renewable energy sources (RESs), load classification strategies, system constraints, dynamic pricing mechanisms, user behavior categorization methods, and various optimization algorithms. Lilliu, F. et al. [47] focused on the commercial application models of energy flexibility at the individual building level. Through a systematic review, they proposed market-oriented implementation pathways for realizing the value of building-level flexibility, with emphasis on innovative business models and dynamic pricing design methods, thus providing a theoretical framework for energy flexibility transactions. From a fundamental theoretical perspective, Cai, S. et al. [48] established a definition and classification system for energy flexibility in individual buildings. They thoroughly analyzed the key enabling technologies for flexibility generation, such as intelligent electrical systems and energy storage integration, and identified four typical application scenarios for building-level flexibility: Positive Energy Districts (PEDs), flexible communities, virtual power plants (VPPs), and integrated energy systems for transportation networks. These scenarios form a technical roadmap for the large-scale aggregation and application of building flexibility resources.

1.3. The Shortcomings of Existing Reviews and the Innovation of This Review

1.3.1. Limitations of Existing Research

Existing review studies on building flexibility resources exhibit three major limitations. First, in terms of classification dimensions, current research often focuses on specific building types (e.g., residential buildings), lacking systematic comparative analyses across different functional categories such as commercial buildings, public buildings, residential complexes, and industrial facilities. This results in an insufficient understanding of the differences in flexibility characteristics, response potential, and technological pathways across building types. Second, regarding resource coverage, most studies concentrate on individual typical resources such as EVs and air conditioning loads, failing to comprehensively identify the coordination potential and complementary mechanisms among diverse flexible resources. Third, at the spatial scale, research is primarily limited to individual building-level optimization, with a lack of systematic exploration of multi-scale resources, including individual buildings, building clusters, and regional-level systems.

1.3.2. Innovations of This Study

This study focuses on energy systems at the regional scale. By establishing a hierarchical classification framework for three types of flexible resources—distributed energy supply, energy storage, and flexible loads—it systematically analyzes the flexibility characteristics, technical constraints, and optimization pathways of typical flexible resources. Furthermore, it reveals the synergistic effects and coupling mechanisms among different types of flexible resources. The research findings provide theoretical support and methodological guidance for addressing the operational challenges of regional energy systems under the high penetration of renewable energy sources.

2. Distributed Energy Supply

2.1. Distributed Renewable Energy Sources

Distributed renewable energy generation technologies applied in the building sector are primarily based on photovoltaic (PV) and wind power systems. These generation systems exhibit dual operational characteristics: while meeting the electricity demand of the building itself, they can also export excess electricity to the grid when generation exceeds onsite consumption, thereby endowing buildings with the dual role of both energy producers and consumers. However, due to the time-varying nature of meteorological parameters such as temperature, solar irradiance (affecting PV), wind speed, and air density, the output power of such systems is characterized by significant intermittency and volatility [49]. Due to adverse weather conditions or seasonal variations, wind and solar power may experience prolonged periods of extremely low output, lasting for hours or even days. This phenomenon, known as a “Dunkelflaute” event [50], poses a significant threat to the reliable operation and economic efficiency of power systems. Furthermore, research indicates that even under balanced load conditions, the high penetration of distributed renewable energy can exacerbate issues within low-voltage distribution networks, including imbalances in voltage/current magnitude and phase angle [51]. These issues can lead to severe overloading and increased neutral line losses, which in critical situations may trigger major incidents such as fire hazards [52] and system outages [53]. These multifaceted challenges—inherent output variability, extreme energy scarcity events, and induced grid instability—render distributed renewable energy sources incapable of directly responding to grid dispatch signals. Consequently, they are classified as non-dispatchable resources that must be coordinated with other flexible assets to effectively contribute to grid regulation.

2.1.1. Coupling Characteristics of Distributed Renewable Energy and Buildings

The flexibility regulation capability of distributed renewable energy is essentially determined by its coupling relationship with building loads. As shown in Figure 1, this multi-dimensional coupling can be analyzed from four key perspectives: temporal, capacity, spatial, and grid interaction dimensions. We have detailedly listed the main coupling characteristics between distributed renewable energy and building loads in

Table 1. Coupling Characteristics between Distributed Renewable Energy and Building Loads.

Type of Relationship		Core Characteristics	Challenges and Solutions
Temporal Dimension	Synchronous Matching	Generation–load time alignment with high self-consumption	No need for storage, but accurate forecasting of load and generation curves is required
	Asynchronous Mismatch	Reliance on storage or grid, with curtailment risk	Deploy energy storage and demand response strategies
Capacity Dimension	Supply	Surplus generation over demand, requiring excess energy handling	Optimize installed capacity; implement dynamic pricing and power trading mechanisms
	Surplus	Strong dependence on external sources	Enable multi-energy complementation (e.g., PV–storage–diesel integration)
Spatial Dimension	Local Direct Supply	Point-to-point energy supply to building loads	Limited by physical constraints of buildings
	Regional Interconnection	Multi-node energy sharing across regions	Apply intelligent dispatch algorithms to balance local supply–demand differences
Grid Interaction	Grid-connected Complementarity	Bidirectional interface with the main grid	Must comply with grid access requirements and dispatch protocols
	Off-grid Operation	High reliability through redundant system design	Configure hybrid energy systems and reserve backup power

, and visualized them in Figure 1.

Temporal coupling characteristics can be divided into synchronous matching and asynchronous mismatch. When the peak output of renewable energy generation coincides with the peak demand of building loads, a high self-consumption rate is achieved (e.g., the synergy between photovoltaic systems and daytime air conditioning loads in commercial buildings) [54]. In such scenarios, building loads can directly utilize locally generated renewable energy, thereby reducing reliance on energy storage or the grid and significantly improving system economics and energy efficiency. In contrast, an asynchronous mismatch occurs when there is a temporal offset between energy generation and consumption (e.g., insufficient electricity supply from residential PV systems during nighttime), requiring energy storage for peak shifting or grid interaction to avoid high curtailment rates of solar or wind energy [55].

Capacity coupling characteristics can be classified into supply surplus and supply deficit conditions. In a supply surplus scenario, the output of renewable energy exceeds the building load demand, posing challenges related to the economic viability of surplus power storage [56] and the coordination of reverse power flow impacts [57]. Conversely, under supply deficit conditions—when renewable energy output is insufficient (e.g., during overcast weather or extreme load demand)—it is essential to ensure energy reliability through multi-energy complementation, such as the combined use of gas turbines and energy storage systems. This is particularly critical for energy-intensive buildings or those with sensitive loads, such as data centers.

Spatial coupling characteristics encompass local direct supply and regional interconnection. The local direct supply model refers to point-to-point energy delivery from distributed energy sources (e.g., rooftop photovoltaics or ground-source heat pumps) to building loads, offering advantages of low transmission losses and high energy efficiency. However, its application is constrained by physical building conditions, such as the available rooftop area for PV installation and the distribution of geothermal resources. The regional interconnection model involves multi-node energy sharing through microgrids (e.g., community-level PV–storage–building load clusters), which requires intelligent dispatch algorithms to optimize regional energy flow and balance local supply–demand disparities.

Grid interaction characteristics can be categorized into grid-connected and off-grid modes. Grid-connected systems interact with the main grid through bidirectional energy interfaces, enabling dynamic balancing via the feed-in of surplus electricity and grid purchasing during shortages. Such systems must strictly comply with grid dispatch regulations, including voltage and frequency fluctuation limits [58], as well as technical standards for grid connection. Off-grid systems rely on local energy autonomy and ensure supply reliability through multi-energy redundancy designs (e.g., photovoltaic systems combined with energy storage and diesel generators) [59], making them suitable for special scenarios such as microgrids in remote areas [60] or emergency power supply.

2.1.2. Flexible Regulation Strategies and Core Challenges

The primary modes of flexible regulation for distributed renewable energy include real-time power matching and time-shifted energy consumption. Among them, real-time power matching focuses on dynamically regulating energy storage systems and flexible loads to achieve immediate balance between generation and demand, thereby mitigating fluctuations and disturbances to the power grid [61,62]. Typical application scenarios include regional microgrids (either off-grid or grid-connected) that require the real-time balancing of generation and load, as well as high-penetration distributed energy distribution networks (e.g., PV-intensive urban grids where reverse power flow may cause voltage violations). The key to achieving real-time power matching lies in the “forecast–control–coordination” technology chain, enabling the instantaneous coupling of generation and load. Representative technical pathways include multi-source data-driven hybrid forecasting models, model predictive control (MPC) frameworks, and hierarchical coordinated control architectures. Hybrid forecasting leverages diverse data sources (e.g., CNN-LSTM fusion using meteorological inputs to enhance PV forecasting accuracy [63], or physics–machine learning hybrid models to optimize load forecasting [64]) to mitigate accumulated prediction errors caused by sudden weather changes or abnormal load behavior [65]. Model predictive control, combined with adaptive algorithms (e.g., Q-learning for dynamic adjustment of storage strategies [66], or reserving backup capacity to address uncertainties), enables second-level response speed and robustness, addressing issues such as device response latency and the economic limitations of frequent adjustments [67,68]. A hierarchical control architecture (e.g., comprising master control–energy management system (EMS) device coordination layers [69]) along with multi-device complementarity technologies, helps overcome coordination efficiency bottlenecks between energy storage and flexible load resources [70,71]. The main bottlenecks facing real-time power matching include prediction error accumulation, device response delay, and economic constraints. Prediction error accumulation arises when weather variability or sudden load anomalies render control strategies ineffective. Device response delay involves the reduced charge/discharge efficiency of storage systems and inverter communication latency. Economic constraints relate to the accelerated degradation of storage systems due to frequent regulation, increasing overall operational and maintenance costs.

The time-shifted energy consumption mode of distributed renewable energy refers to a systemic solution aimed at dynamically balancing power supply and demand by efficiently utilizing or reallocating surplus electricity during specific periods when distributed renewable generation (such as photovoltaics or distributed wind power) significantly exceeds local real-time demand—for example, at noon on sunny days or during high-wind periods—through technological, managerial, and market-based approaches, thereby reducing wind and solar curtailment [72]. The core mechanism for achieving this mode lies in restructuring the generation–load interaction through a dual strategy of “intertemporal energy transfer via storage + proactive adaptation of flexible loads.” On the load side, technologies such as interruptible industrial load control strategies [73,74], the utilization of the thermal inertia of building HVAC systems [75,76], and the intelligent charging/discharging control of EVs [77,78] enable dynamic alignment between load profiles and the fluctuating output of renewable energy. Meanwhile, leveraging the intertemporal energy storage capabilities of energy storage systems within a multi-energy complementary operation framework transforms intermittent surplus renewable power into dispatchable and spatially adaptable flexibility resources [79,80]. This mechanism establishes a bidirectional, dynamic balance between supply and demand through coordinated interaction between the demand side and the supply side, thereby significantly improving system resilience to renewable energy variability. The main challenges of time-shifted consumption include limited energy storage capacity, high cost pressure, and the complexity of coordinating heterogeneous resources such as HVAC systems and EVs.

In summary, real-time power matching focuses on dynamic balancing at second- to minute-level timescales, with its core value lying in mitigating the impact of stochastic fluctuations from distributed energy on the power grid. In contrast, time-shifted energy consumption targets energy redistribution across hour- to day-level timescales, aiming to enhance the overall absorption capacity of renewable energy. A schematic diagram of these modes is presented in Figure 2. The two approaches are complementary in terms of temporal resolution, regulation objectives, and technical pathways: the former ensures real-time grid stability through fine-grained control, while the latter improves system economic efficiency through large-scale scheduling. Together, they form a multi-timescale flexible regulation framework for distributed energy systems.

2.2. Building-Integrated Multi-Energy Cogeneration Technologies

Building-integrated multi-energy cogeneration technologies achieve efficient energy conversion and supply through energy cascade utilization and multi-energy synergy. The core concept involves integrating originally separate systems for power generation, heating, and cooling into a unified system that recovers waste heat generated during power production (e.g., exhaust heat from gas turbines or internal combustion engines) to drive heat exchangers, absorption chillers, and other equipment, thereby simultaneously meeting the electricity, heating, and cooling demands of buildings and significantly improving primary energy utilization efficiency (typically saving 10–30% compared to traditional separate generation systems) [81]. The main forms include Combined Heat and Power (CHP) and Combined Cooling, Heating, and Power (CCHP). CHP refers to the simultaneous production of electricity and heat without cooling capability and is particularly suitable for heat-dominated buildings, such as the heating demand for buildings in northern China [82]. CCHP extends the CHP system by integrating cooling functions, thereby enabling the simultaneous supply of electricity, heat, and cooling. It is especially applicable to regions with balanced heating and cooling demands, such as the commercial building areas in the southeast of China.

2.2.1. Flexibility Characteristics and Strategies

CCHP and CHP systems can respond to grid requirements for peak shaving, frequency regulation, reserve capacity, renewable energy integration, and demand-side management through coordinated multi-energy flow conversion, thermal storage technologies, and intelligent control strategies. The operating principles of these systems are illustrated in Figure 3. In terms of peak shaving, these systems dynamically adjust the output of electricity, heating, and cooling to flatten load curves; intra-day balancing is achieved by storing thermal or cooling energy during off-peak electricity price periods and reducing grid power purchases during peak hours [83]. Seasonal peak regulation is realized by adjusting the cogeneration ratio based on thermal demand characteristics—prioritizing heating in winter to alleviate grid stress during cold peaks, and using cooling storage systems in summer to balance cooling loads with electricity demand. For frequency regulation, gas turbine-driven units can participate in second- to minute-level grid frequency control thanks to their rapid ramping capabilities, while battery energy storage systems (BESSs) provide voltage and frequency support [84]. In renewable energy integration scenarios, power-to-heat technologies convert surplus wind or solar electricity into thermal energy for storage, or use thermal storage to smooth intermittent renewable output. Demand-side response is enabled through coordinated control of electrical, thermal, and cooling loads, allowing flexible time shifting, such as dynamically prioritizing air conditioning or industrial heat loads. Additionally, these systems can support islanded operation and black start capabilities.

CCHP and CHP systems primarily contribute to grid flexibility through parameter regulation and strategic control. In terms of parameter regulation, both systems dynamically adjust the operating parameters of key equipment—such as gas turbine load, steam flow rate in absorption chillers, and thermal output power—to achieve coordinated multi-energy optimization across electricity, heat, and (for CCHP) cooling. They also leverage the storage capacity and charge/discharge rates of various storage systems (e.g., batteries, hot water tanks, and ice storage units) to mitigate grid fluctuations and respond to real-time energy demands. Strategically, both systems adopt hierarchical control frameworks: at the top level, economic models (e.g., real-time pricing response and ancillary service bidding) are used to optimize global operational objectives; at the middle level, MPC or rolling optimization algorithms forecast load and price trends to adjust multi-energy output ratios; and, at the bottom level, PID controllers or fuzzy logic are used for rapid response to transient grid changes in frequency and voltage. Due to its integrated supply of electricity, heating, and cooling, CCHP offers broader regulation capabilities through flexible switching between output modes (e.g., prioritizing power generation or heat supply), while CHP focuses on enhancing electrical regulation independence via the coupling characteristics of thermal and electrical loads (e.g., through thermal–electric decoupling technologies). Both systems are deeply embedded in the flexible grid regulation ecosystem through participation in demand response programs, reserve capacity markets, and black start services.

2.2.2. Core Challenges and Solution Pathways

The core challenges of building-integrated cogeneration technologies in participating in grid flexibility regulation stem primarily from the strong coupling between thermal and electrical outputs, which limits operational flexibility. In conventional CHP systems, power output is often constrained by thermal demand (heat-led operation), making it difficult to respond independently to grid power regulation requirements. For instance, the waste heat generated by gas turbines must first meet heating needs, resulting in rigid power output and restricted capacity to participate in peak shaving or frequency regulation [85,86]. In addition, thermal systems exhibit significant inertia and slow dynamic response, which creates a mismatch with the fast-response requirements of grid regulation, thereby limiting real-time adjustability [86,87]. The complexity of controlling multi-energy systems—such as coordinating the operation of generation, heating, thermal storage, and electrical storage subsystems—poses further challenges in the form of dynamic multi-variable optimization problems [88]. A fundamental trade-off also exists between system efficiency and flexibility: improving flexibility through rapid start-up/shutdown or partial-load operation often reduces device efficiency and increases operational costs, compromising economic viability [89]. Moreover, most existing control architectures lack coordinated optimization mechanisms that consider thermal–electric coupling constraints, with traditional “heat-led” or “power-led” control modes falling short of the flexibility and robustness required for dynamic grid participation [86,90]. To address these issues, several solution pathways have been proposed: (1) physical decoupling technologies, such as large-capacity thermal storage units (e.g., high-temperature thermal tanks) or devices with adjustable heat-to-power ratios (e.g., micro gas turbines), can buffer surplus heat or flexibly modulate output to decouple heat and power flows; (2) optimized control strategies, including hierarchical control architectures and MPC, can dynamically balance thermal–electric supply and grid demands; (3) multi-energy hybrid systems that integrate renewable sources (e.g., solar PV, wind) with joint electrical and thermal storage enhance overall system flexibility; and (4) economic and policy instruments, such as demand response programs and time-of-use pricing incentives, can further support flexible system operation.

While the capacity of building-integrated multi-energy cogeneration technologies is relatively small, necessitating aggregation for participation in demand response programs, expanding the perspective to a regional level reveals that District Heating Networks (DHNs) are natural aggregators of thermal demand and possess substantial flexibility potential. Modern DHNs can integrate large-scale, centralized power-to-heat (P2H) assets, such as electric boilers and district-level heat pumps [91]. Utilizing numerous P2H devices facilitates the absorption of surplus low-cost electricity from variable renewable sources like wind and photovoltaics, converting it into useful thermal energy [92]. This provides an effective mechanism for mitigating the intermittency of renewable energy and offering ancillary services to the upstream electricity grid [93]. Simultaneously, DHNs can be coupled with cost-effective options like centralized thermal energy storage (TES) tanks [94,95], the thermal mass of building envelopes [94], and the water within the heating network itself [96]. This integration enables the decoupling of heat production timing (e.g., during renewable energy generation peaks) from consumption timing, thereby providing a significant flexibility resource [97]. Consequently, the holistic and coordinated optimization of integrated electricity and district heating systems represents a critical frontier for future research.

2.3. Summary

Distributed renewable energy and building-integrated cogeneration technologies, as core flexible resources in distributed energy supply systems, exhibit complementary characteristics in terms of technical features and functional positioning. Distributed renewable energy is primarily characterized by intermittent output, requiring coordination with energy storage systems and the power grid to achieve power balance and energy shifting across time scales ranging from seconds to days. Its core technical bottlenecks lie in output fluctuations that pose risks to grid stability and the cost constraints associated with the full life cycle of energy storage systems. In contrast, building-integrated cogeneration technologies establish multi-energy supply systems of electricity, heat, and cooling based on the principle of energy cascade utilization. Although their thermoelectric coupling limits independent regulation on the electricity side, they can still effectively participate in ancillary services such as peak shaving and frequency regulation through the optimal configuration of thermal/cooling storage units and the fine-tuning of operational parameters. The two types of technologies complement each other through synergistic integration: distributed renewables provide low-carbon electricity input to the cogeneration systems, while cogeneration technologies enhance system flexibility through thermoelectric decoupling and coordinated multi-energy flow optimization. The resulting integrated energy supply architecture—combining renewable energy, multi-energy cogeneration, and energy storage—can simultaneously improve comprehensive energy utilization efficiency on the building side and meet the flexible regulation needs of the power grid across multiple time scales.

3. Distributed Energy Storage

Energy storage serves as a pivotal flexible component in regional energy systems, with its essential function being to act as a temporal buffer that decouples energy supply from demand through the storage and release of energy over time [98]. Energy storage systems (ESSs) possess bidirectional power flow capabilities, enabling them to absorb energy from the grid or renewable sources (acting as a load) and inject it back into the grid (acting as a generation source). This bidirectionality allows ESSs to provide solutions for mitigating power output fluctuations, maintaining frequency, and ensuring voltage stability in smart grids [99].

Based on the form in which energy is stored, ESSs can be classified by their storage medium into categories such as mechanical, electrochemical, chemical, and thermal [100]. The key technical characteristics, pros and cons for flexibility applications, primary challenges and limitations, and typical application scenarios for these different energy storage forms are summarized in

Table 2. Comparison of Major Energy Storage Technologies.

Type	Device	Definition	Characteristics	Main Challenges	Application	Ref.
Electrical	Super- Capacitor	A device that stores energy via electrostatic charge accumulation at the interface between an electrode and an electrolyte.	High charge/discharge rates, high efficiency, extended lifespan, and durability.	Low specific energy, which limits their capability for continuous power supply, and a high self-discharge rate.	Supporting wind/photovoltaic grid integration; responding to rapid power fluctuations.	[101,102]
Thermal	Sensible Heat Storage	Storing or releasing thermal energy by changing the temperature of a storage medium.	Mature technology, simple and reliable, low-cost, and flexible system design.	Low energy storage density, degradation of energy quality (exergy) due to temperature variations, and significant heat loss.	Assists in enhancing the energy flexibility of district heating, building air conditioning, and industrial production processes	[103,104,105]
	Latent Heat Storage	Utilizing the absorption or release of thermal energy from a phase change material (PCM) during its phase transition.	High volumetric heat storage density and stable charge/discharge temperatures.	Relatively high cost, low thermal conductivity, slow response speed, and poor stability.	Providing energy flexibility support for building-level HVAC systems.	[106,107]
	Thermochemical Heat Storage	Storing energy through the endothermic and exothermic properties of reversible chemical reactions.	Highest energy density and low storage losses.	Technological immaturity and high cost.	Cross-seasonal energy storage and high-grade thermal energy storage.	[108,109]
Electro-Chemical	Lithium Ion Battery	Achieves charging and discharging through the intercalation and de-intercalation of lithium ions between the cathode and anode.	Fast response time, high energy density, high efficiency, and negligible memory effect.	High initial cost and poor thermal safety are the primary limitations.	Commonly used at the regional level to facilitate large-scale renewable energy accommodation and participate in grid dispatch.	[110,111,112]
	Flow Battery	Active materials in the electrolyte solution at the positive and negative electrodes undergo reversible redox reactions to achieve the interconversion of electrical and chemical energy.	High safety, high power rating, long cycle life, and fast response speed.	Low energy density, system complexity, and relatively low round-trip efficiency.	Integration with distributed energy resources (DERs); smoothing the output of renewable energy sources.	[113,114]
	Sodium-sulfur Battery (NaS)	A high-temperature battery that stores and releases electrical energy through the conduction of sodium ions.	Balanced power capability and energy density, fast response speed, and favorable long-term economics.	Stringent operating conditions, high safety requirements, and limited application scenarios.	Suitable for large-scale, stationary energy storage applications.	[115,116]
Chemical	Hydrogen-based Energy Storage	Stores electrical energy by producing hydrogen via an electrolyzer; the stored hydrogen is then used in a fuel cell to generate electricity.	High efficiency, clean, safe, environmentally friendly, and high energy density.	High cost of electrolysis; complex storage and transportation challenges.	Flexible applications for both short-duration and long-term/seasonal energy storage.	[117,118]
Mechanical	FlyWheel	Stores electrical energy by converting it into the kinetic energy of a rotor.	Fast response speed, extremely long cycle life, and high charge/discharge rates.	Very short storage duration, high self-discharge rate (standby losses), and high cost.	Typically integrated with industrial loads or large-scale equipment like thermal power units.	[119,120]

.

Although a systematic comparison of various energy storage technologies is provided in the preceding section and in Table 2, their applicability and deployment scale differ markedly within the practical context of regional energy systems. Among these technologies, electrical energy storage (EES) and thermal energy storage (TES) have emerged as pivotal solutions for enhancing system flexibility, owing to their inherent operational agility and strong coupling characteristics with regional energy infrastructures. Consequently, the subsequent discussion will focus specifically on these two storage modalities. It will delve into their concrete applications and synergistic potential for integrating multi-vector energy demands—namely cooling, heating, and electricity—and for facilitating grid-friendly interactions.

3.1. Electrical Energy Storage (EES)

Electrical energy storage technologies function by converting electricity into other forms of energy—such as chemical, mechanical, or thermal—for storage, and reconverting it back into electricity when needed [121,122]. These technologies employ diverse mechanisms to meet the specific demands of the building sector. Typical applications include time-of-use arbitrage, renewable energy integration, backup power supply, grid regulation, and coordination with CHP systems. Integration modes can be categorized into distributed systems (e.g., residential units or commercial buildings) and centralized systems (e.g., microgrids or CHP-based infrastructures), and can operate under grid-connected, off-grid, or hybrid configurations for enhanced flexibility.

3.1.1. Flexibility Characteristics and Strategies

The flexible regulation characteristics of electrical energy storage systems—including fast response, bidirectional control, and multi-timescale adaptability—make them a key enabler for enhancing grid flexibility [123]. Figure 4 provides a visual summary of these characteristics. First, they offer rapid power regulation capabilities: chemical storage systems (e.g., lithium-ion batteries) can respond within milliseconds to seconds, making them suitable for frequency regulation, while electromagnetic storage systems (e.g., supercapacitors) achieve sub-millisecond responses, ideal for managing instantaneous power fluctuations. These systems also exhibit power/capacity scalability, as their rated power and storage capacity can be adjusted through series or parallel configurations to meet the needs of various grid scales. Secondly, they support bidirectional energy flow, enabling seamless switching between charging and discharging modes based on grid conditions. Charging can absorb excess electricity—such as from renewable generation surpluses or off-peak periods—while discharging can supply power during peak loads or frequency drops. Furthermore, they enable multi-timescale control: short-term response (on the scale of seconds to minutes) addresses transient issues like frequency deviations and voltage sags, while medium- to long-term modulation (from hours to days) facilitates peak shaving, renewable energy integration, and demand-side response.

The flexible regulation capabilities of EES in the building sector are primarily reflected in two dimensions. First, ESS can mitigate the output fluctuations of distributed renewable energy sources by absorbing and releasing energy, thereby reducing the impact of intermittent sources such as photovoltaics on the grid. Second, EES can dynamically adjust charging and discharging strategies based on real-time dispatch signals, enhancing the operational flexibility of building energy systems. For example, Gupta, R. et al. [124] deployed PV and EES systems across 82 residential clusters, demonstrating that the integration significantly improved local photovoltaic self-consumption and reduced peak-period electricity demand by an average of 8%. Similarly, research by Niu, J. et al. [125] showed that the use of EES can substantially enhance a building’s flexibility, and leveraging time-of-use electricity pricing can reduce operational costs by 5.3%.

3.1.2. Core Challenges and Solution Pathways

Although building-integrated EES offers significant technical advantages in supporting grid flexibility, they still face multiple critical constraints that must be addressed. First, battery cycle life and performance degradation pose direct challenges to long-term system sustainability. Frequent charge–discharge cycles—common in applications such as frequency regulation and fast-response services—accelerate battery aging, leading to capacity fade and efficiency losses [126]. Research indicates that a 12% increase in EES flexibility intensity may result in a 20% reduction in battery lifespan [11]. Second, high upfront investment and operational costs constitute major economic barriers. These include ongoing expenditures for battery replacement, thermal management, and monitoring systems, making the payback period generally exceed 10 years in regions with narrow electricity price differentials or underdeveloped ancillary service markets [127]. Third, the inherent trade-off between response speed and energy density limits the regulation capabilities of different storage technologies. High-speed devices such as supercapacitors [102,128] and flywheels [129,130] can respond within milliseconds—suitable for high-frequency, short-duration services—but suffer from low energy density. Conversely, lithium-ion [131] and flow batteries [132,133], which offer higher energy densities, require several seconds to minutes for response, making them less ideal for instant grid fluctuations. This diversity in technical characteristics necessitates scenario-specific technology selection, as no single EES type can fulfill all flexibility requirements. Lastly, certain storage technologies raise environmental and ethical concerns. For example, lead–acid batteries contain heavy metals such as lead and cadmium, while flow batteries may involve toxic bromine compounds, both of which present leakage risks that can threaten ecosystems and human health [72]. These environmental and ethical issues significantly dampen public willingness to participate in demand response programs [134].

3.2. Thermal Energy Storage (TES)

Thermal energy storage (TES) technologies can be categorized into two main pathways: passive building thermal mass and active thermal storage systems. Building thermal mass refers to passive thermal capacity systems consisting of envelope elements (e.g., external walls, roofs) and interior components (e.g., furniture, internal walls) that inherently store and release heat. The heat exchange process in such systems is driven by the temperature difference between the environment and the surface of thermal mass materials [135,136]. In contrast, active thermal storage systems integrate dedicated thermal storage devices and modules into energy systems to enable controllable thermal energy dispatch. These primarily include sensible heat storage and latent heat storage. Sensible heat storage stores thermal energy by changing the temperature of a storage medium (e.g., water or rock), while latent heat storage relies on phase change materials (PCMs) to absorb and release heat during material phase transitions [98,137].

3.2.1. Flexibility Characteristics and Strategies

The flexibility of building thermal mass is primarily reflected in its ability to respond to environmental or load fluctuations, characterized by both response speed and regulation capacity. Fast thermal responses can be achieved using PCM or high-conductivity materials that allow rapid heat absorption and release, while slow responses rely on materials with high thermal capacity, such as concrete or water, to enable long-duration thermal storage. The synergy between the thermal storage capacity of building mass and pre-cooling strategies offers an innovative solution for shifting cooling loads across time and space and smoothing building electricity demand peaks and troughs. Turner, W.J.N. et al. [138] quantitatively analyzed this synergistic approach and found that 50% to 99% of peak cooling electricity demand could be shifted to off-peak hours annually. However, conventional building thermal mass is constrained by its inherent thermal inertia, limiting its ability to meet high-flexibility requirements. In this context, the innovative application of PCM presents a promising avenue for overcoming these limitations. Dong, Y. et al. [9] investigated the integration of PCM into building envelopes in office buildings and found that when combined with HVAC pre-cooling strategies, PCM-enhanced walls improved load-shifting capability by 69.7% and reduced total energy load by 1.3% compared to conventional walls, achieving a balance between flexibility and energy efficiency. Johra, H. et al. [136] demonstrated that embedding PCM into indoor furniture significantly enhanced a building’s energy regulation capacity. Their experiments revealed that this approach increased the energy flexibility index by 87% and 30% in low-insulation and high-insulation lightweight structures, respectively. Similarly, Le Dréau, J. et al. [139] showed that the flexibility potential of building thermal mass is influenced by insulation levels and HVAC terminal types. Reynders, G. et al. [140] further proposed a quantitative model for building thermal mass flexibility, revealing the coupled effects of insulation performance and dynamic boundary conditions on load-shifting potential. Collectively, these findings highlight how the interdisciplinary convergence of materials science and building thermal engineering is driving the development of intelligent thermal responsiveness. Through innovations in advanced thermal storage materials and integrated envelope design, the traditional limits of passive thermal mass can be overcome, expanding the potential for flexible energy regulation in buildings.

The coordinated coupling of thermal energy storage (TES) systems with thermoelectric conversion devices—such as heat pumps—represents a critical technical pathway for improving both the efficiency and flexibility of integrated energy systems. Patteeuw, D. et al. [141] combined electric heating systems for domestic hot water and space heating with thermal storage tanks and building thermal mass to enable active demand response in residential buildings, resulting in an overall operational cost reduction of approximately 1.8% on the demand side. Baeten, B. et al. [142] developed a coordinated heating system integrating heat pumps with hot water storage tanks, demonstrating that compared to standalone heat pump systems, the inclusion of thermal storage reduced power plant energy intensity during peak load periods by 11%. Su, F. et al. [143] evaluated the performance of integrating ice storage systems into HVAC configurations in commercial buildings (e.g., shopping malls), and across 11 cities with different climate zones, time-of-use pricing combined with ice storage led to electricity bill savings ranging from 17.37% to 44.92%, showcasing significant peak shaving effects. Liu, X. et al. [144] proposed a single heat pump system coupled with a thermal storage tank and found that by optimizing control strategies, peak load could be reduced by up to 60.8%. Hu, Y. et al. [145] designed a PCM thermal storage unit integrated with a heat pump system, and under time-of-use electricity pricing, the system achieved a 7% reduction in energy costs compared to conventional HVAC systems without PCM integration.

3.2.2. Core Challenges and Solution Pathways

Building thermal energy storage technologies demonstrate strong potential for peak shaving and balancing the intermittency of renewable energy sources within the context of grid flexibility. However, their practical application faces significant technical and economic challenges.

First, in terms of temporal response characteristics, the high thermal inertia and unidirectional regulation nature of TES systems limit their ability to participate in high-frequency grid regulation. Thermal inertia arises from the limited heat conduction rates of storage materials, resulting in charge/discharge processes that typically respond on the scale of tens of minutes to several hours [146,147], which falls short of the seconds-to-minutes response times required for grid frequency regulation. Moreover, most TES systems only support unidirectional regulation—for example, discharging heat to offset power deficits or absorbing surplus electricity through heat storage—without the bidirectional flexibility found in electrochemical storage (e.g., lithium-ion batteries) or mechanical storage (e.g., flywheels). This restricts TES applications to long-duration regulation scenarios, such as intra-day peak load shifting, while limiting their effectiveness in rapid-response or ancillary service markets [142,148,149].

Second, the strong coupling between power output and storage capacity imposes a significant limitation on TES performance. The maximum thermal power output of a storage device is directly constrained by its current thermal energy level—for example, the remaining heat content in a thermal tank directly affects its discharge rate—thereby preventing independent adjustment of power and capacity [150]. To address this issue, modular system design has emerged as a key optimization strategy. By deploying distributed clusters of small-scale thermal storage units (e.g., prefabricated phase-change storage modules operated in parallel), it is possible to dynamically scale output and allocate power precisely based on demand, effectively decoupling the physical dependency between capacity and power [151].

Additionally, economic and market mechanism barriers stem from the limited service valuation and revenue models available to TES systems. Currently, electricity markets tend to offer lower compensation for long-duration peak-shaving services compared to short-duration frequency regulation, thereby underrepresenting the economic potential of TES. Moreover, the financial viability of TES is often dependent on long-term electricity price arbitrage [149,152], which is highly sensitive to market volatility and policy stability—factors that collectively extend the investment payback period [153]. To address these challenges, market structures and business models must be redefined. Implementing time- and location-differentiated pricing mechanisms can more accurately reflect the value of grid flexibility services and establish differentiated compensation schemes for long-duration and short-duration applications, thereby enhancing TES economic viability across multiple regulation scenarios. In parallel, the adoption of capacity leasing models should be promoted, allowing third-party operators to aggregate distributed building-based TES resources. Under a “storage-as-a-service” framework [154,155], these operators can participate in electricity markets on behalf of end-users, alleviating the upfront investment burden and improving the scalability and feasibility of TES deployment [156].

3.3. Summary

The core of distributed energy storage technologies lies in the coordinated integration of EES and TES, which exhibit complementary advantages in both technical performance and economic feasibility [157]. EES offers high energy density and excellent power response capabilities, enabling direct participation in grid flexibility services and delivering high-quality instantaneous power support. In contrast, TES, despite its slower dynamic response, stands out in terms of cost-effectiveness, making it well-suited for energy shifting applications. Therefore, integrating the differentiated characteristics of EES and TES to construct a flexible building energy regulation system that balances rapid responsiveness with economic viability has become a central research focus in energy storage integration. For example, in PV–heat pump hybrid systems [158], the coordinated optimization of electrical and thermal storage significantly enhances load management capabilities. At the same time, leveraging the economic and environmentally compatible features of TES facilitates multi-objective optimization across technical, economic, and environmental dimensions, offering a new technological paradigm for building–grid interaction.

4. Flexible Load

4.1. Air Conditioning Load

Air conditioning load is the most promising flexibility resource among building loads, with its core advantages reflected in three aspects: (1) a significant proportion of energy consumption and ample controllable capacity; (2) the thermal comfort range exhibits elasticity, which helps reduce the impact on user satisfaction; and (3) the ease of integration with intelligent energy management systems and the simplicity of control operations [159]. The flexibility regulation strategies for air conditioning loads mainly depend on the type of air conditioning system. For split-type air conditioners, flexible control is primarily achieved through indoor temperature setpoint adjustment and pre-cooling/pre-heating strategies. Centralized air conditioning systems involve more complex control dimensions, including indoor temperature setpoint adjustment, air flow distribution control, and chiller unit regulation.

4.1.1. Flexibility Characteristics and Strategies

The main flexible regulation strategies for different air conditioning loads are illustrated in Figure 5.

Indoor Temperature Setpoint Adjustment: Indoor temperature setpoint adjustment mainly involves proactively regulating the building’s cooling/heating load curve by dynamically modifying the temperature control parameters (±2 °C) of the air conditioning system within the thermal comfort range. As a typical form of direct load control, this strategy can provide the power grid with a rapidly responsive adjustable load resource. Its technical features include: an initial response capability within minutes, a ramp response time of 5–30 min, and the ability to sustain load regulation for 0.5–4 h [36]. Existing studies have shown that moderate adjustments to the temperature setpoint can significantly reduce building peak energy consumption. Chen, Y. et al. [160], through a systematic quantitative analysis, demonstrated that increasing the HVAC temperature setpoint in office buildings by 2 °C (from 24 °C to 26 °C) during demand response periods can achieve a peak load reduction of 1.9 kW. Triolo, R.C. et al. [7] showed that implementing a 1.1 °C (2 °F) temperature setpoint increase across a cluster of 124 buildings can reduce peak cooling load on the regional cooling system by over 10% on the highest load day of the year. However, this strategy requires attention to potential load rebound effects during implementation—i.e., the increased air conditioning load required to restore indoor thermal conditions after the DR event may cause a new power peak. To address this challenge, researchers have proposed various solutions. Meng, Q. et al. [10] studied the effects of control strategies such as zone temperature reset, active thermal storage, the activation/deactivation of cooling sources, and pre-cooling on the flexibility of HVAC systems and the mitigation of rebound effects. Their results indicate that integrating a thermal storage tank into the HVAC system can reduce rebound power by approximately 3%, and extending the DR duration to the building’s thermal inertia threshold can completely eliminate the rebound effect. Talebi, A. et al. [161] developed a 24 h predictive control algorithm that integrates weather and load forecast data to optimize air conditioning operation, achieving a 5% reduction in rebound power.

Pre-cooling/pre-heating: The core concept of the pre-cooling/preheating strategy is to utilize the thermal inertia of the building envelope. By pre-adjusting cooling/heating loads during off-peak electricity price periods, energy is stored in the building’s thermal mass, thereby reducing equipment operation during grid peak hours and achieving temporal and spatial load shifting [162]. Technical studies indicate that the typical load regulation duration for this strategy ranges from 0.5 to 3.0 h, although specific response times and power modulation rates require further quantitative analysis [36]. Optimizing the timing window for implementing the strategy is critical to its effectiveness: overly long pre-conditioning periods may lead to thermal losses through the envelope, reducing energy utilization efficiency, while periods that are too short may not provide effective load flexibility during DR events. One study [163], using rule-based methods to simulate the flexibility potential of the pre-cooling strategy, showed that chiller operation could be halted for more than 90 min during peak periods. Other studies have improved the flexibility potential of the strategy through optimized control methods. For example, one study [164] designed an MPC method based on an R-C model for the pre-cooling control of air conditioning systems, and found that it can reduce energy costs by 20–56.8% compared to rule-based strategies.

Airflow Distribution Control: Central air conditioning systems are characterized by structural complexity and multiple adjustable parameters. Airflow distribution control achieves precise system response regulation by adjusting key parameters such as fan power and static pressure setpoints. This strategy, with its rapid response time (typically <1 min) and short cycle adjustment capability (on the order of seconds to minutes), has become the preferred technical approach for enabling building HVAC systems to participate in grid frequency regulation [36]. Hao, H. et al. [165], through numerical simulations, demonstrated the feasibility of commercial building HVAC fans participating in frequency regulation. Their results showed that approximately 15% of fan rated power can be converted into adjustable capacity without compromising thermal comfort. Zhao, P. et al. [166] further extended the control dimensions of variable air volume (VAV) systems by comparing the performance of static pressure setpoint adjustment and zone temperature setpoint adjustment. They found that the dynamic adjustment of static pressure within a ±125 Pa range (with a control depth of 50%) can provide ±100 kW of instantaneous power regulation capacity.

Chiller Regulation: As the core cooling equipment in central air conditioning systems, chillers account for up to 70% of the system’s total energy consumption, making their operational control a key entry point for flexible building energy management. Current research mainly adopts two typical control modes to implement demand response: (1) coordinated start–stop control of equipment clusters; and (2) dynamic regulation of supply water temperature, achieving cooling load redistribution by modifying the supply return temperature differential. Tang, R. et al. [167] proposed a coordinated control method for start–stop sequences and chilled water flow distribution. By optimizing the startup combinations of chillers and pumps and simultaneously adjusting chilled water distribution, the method achieved a 23% instantaneous power reduction while maintaining zone temperature control accuracy. Moreover, the strategy demonstrated strong rebound suppression capability after DR events, successfully reducing grid peak demand by 3300 kW (34%). Su, L. et al. [168] experimentally explored the dynamic response characteristics of temperature regulation modes. When the chilled water outlet temperature setpoint was dynamically adjusted by ±2.1 °C, the chiller was capable of providing up to ±25% of its rated power as secondary frequency regulation capacity.

4.1.2. Core Challenges and Solution Pathways

Flexible regulation of air conditioning loads faces multidimensional technical challenges, primarily in terms of system dynamic characteristics, multivariable coupled control, and energy efficiency optimization. The core challenge lies in developing a coordinated control model that simultaneously accounts for thermodynamic response characteristics and user comfort requirements, which necessitates addressing the nonlinear relationships among control parameters across multiple time scales. Specifically, the coordinated optimization of temperature setpoint adjustment and pre-cooling/-heating strategies must overcome the delay effects caused by building thermal inertia, while the integrated operation of airflow distribution control and chiller regulation faces the difficulty of balancing localized regulation with overall energy efficiency.

The solution path should focus on constructing a hierarchical control architecture that enables multi-objective optimization by decomposing time constant differences. At the equipment level, MPC algorithms should be employed to manage fast dynamic processes such as fan variable frequency adjustment and terminal flow distribution. At the system level, reinforcement learning algorithms can be used to optimize chiller operation strategies, while integrating building thermal inertia models for load forecasting. Furthermore, a dynamic evaluation system based on human thermal comfort indices should be established, embedding the PMV-PPD model into the control loop to enable flexibility regulation under comfort constraints.

4.2. Electric Vehicles (EVs)

Driven by the global impetus to reduce CO2 emissions and curtail petroleum consumption [169], the worldwide adoption of EVs has surged. However, this rapid proliferation poses significant challenges to electrical distribution networks. For instance, uncoordinated EV charging can create new peaks in load demand [170], high concentrations of charging activities in localized areas may lead to severe voltage fluctuations [171], and the non-linear load characteristics of EV chargers can inject harmonic currents into the grid, causing harmonic distortion and degrading power quality [172,173].

Therefore, the implementation of intelligent and coordinated charging and discharging strategies is not merely an option but a critical imperative. Through synergistic control, large fleets of EVs can be transformed from potential grid burdens into valuable flexible assets capable of providing a range of ancillary services. Consequently, a comprehensive understanding of their integration models, modes of energy exchange, grid service capabilities, and the factors influencing their flexibility is paramount to converting this challenge into a significant opportunity for future energy systems.

4.2.1. Access Modes of Electric Vehicles

As mobile energy storage units, the integration modes of EVs into energy systems can be categorized into three types: building-level access, building cluster-level access, and grid-level comprehensive applications. Figure 6 provides a schematic diagram of these typical access modes.

Building-level access includes Vehicle-to-Home (V2H) and Vehicle-to-Building (V2B). V2H enables bidirectional energy interaction between vehicles and homes through smart charging stations. It serves as an emergency power supply during grid outages and optimizes household electricity costs during peak and off-peak periods. Zafar, B. et al. [174] applied V2H technology to a HEMS to optimize appliance scheduling and EVs charging/discharging time and power, achieving reductions in economic cost and peak load by 33% and 40%, respectively. V2B, applied in commercial/public building scenarios, involves the coordination of vehicle energy storage systems with Building Energy Management Systems (BEMS) to participate in building load regulation and distributed energy consumption. Ding, Y. et al. [175] studied the V2B operation mode integrating EVs with office building air conditioning load systems and PV, and their results showed that coordinated operation of EV smart charging and air conditioning load pre-cooling can achieve 17.57–61.67% load shifting and increase PV utilization efficiency by more than 52.38%. Borge-Diez, D. et al. [176] combined V2H and V2B technologies by charging employee EVs at home during off-peak hours and discharging them at workplaces to supply office buildings, resulting in a 49% reduction in total energy consumption, annual energy savings of €32,882, and a CO₂ emissions reduction of 30,071 kg.

Building cluster-level access includes Vehicle-to-Building-to-Building (V2B²) and Vehicle-to-Community (V2C). V2B² achieves inter-building energy transfer through energy routing technology, enabling dynamic allocation of renewable energy among buildings via EVs, thereby overcoming the energy self-balance limitations of individual buildings. V2C integrates the storage capacity of multiple EVs at the community microgrid level, responding to regional grid ancillary service demands such as frequency regulation and reserve through cluster scheduling algorithms. In building cluster-level access, EVs act as mobile hubs for energy interaction within building clusters, achieving dynamic optimization of cross-building energy allocation through bidirectional energy flows and significantly enhancing the temporal and spatial flexibility of regional energy systems. For example, Barone, G. et al. [177] showed that the V2B² model can achieve the temporal/spatial transfer of distributed energy and energy complementarity among buildings, thus efficiently utilizing renewables and reducing dependence on the power grid. As a result, power purchases from the grid can be reduced by approximately 45–77%.

Grid-level comprehensive applications include Vehicle-to-Grid (V2G) and Vehicle-to-Everything (V2X). V2G aggregates distributed EV storage into a virtual power plant (VPP) through smart metering and communication technologies, enabling direct participation in advanced applications such as grid frequency regulation, peak shaving, and congestion management. V2X extends the interaction capabilities of EVs to various nodes such as charging stations, storage stations, and renewable energy plants, forming a transportation–energy coupled network. Zhang, W. et al. [178] proposed a 5G-V2X coordinated demand response scheduling method, where EVs cooperate with 5G communication base stations to enhance overall energy flexibility. Their results show that under the 5G-V2X framework, coordinated demand response between 5G base stations and EVs can reduce peak/valley load differences by 23.95%. Slama, S.B. et al. [179] integrated V2X (including V2G, V2H, and V2I) with peer-to-peer (P2P) energy trading in industrial load clusters, achieving a 19.18% reduction in energy costs and a 50.02% reduction in average energy prices.

4.2.2. Forms of Energy Interaction in Electric Vehicles

EVs, through the coordinated operation of onboard battery systems and charging infrastructure, simultaneously serve as “mobile energy storage units” and “distributed power generation devices.”

Table 3. Typical EVs Access Modes and Their Application Scenarios.

Mode	Interaction Scale	Core Function	Typical Application Scenario
V2H	Single household	Household energy self-sufficiency	Distributed PV–storage systems
V2B	Individual building	Flexible building load regulation	Commercial complex energy management
V2B2	Building cluster	Cross-building energy routing	Net-zero carbon park development
V2G	Regional grid	Grid ancillary services	Virtual power plant operation

presents the main categories of current EV charging infrastructure, which include standard plug-in charging, conventional slow charging stations, and fast-charging stations [180]. From a technical perspective, standard plug-in and conventional slow-charging stations adopt low-power energy transmission modes. Their lower power ratings and longer charging durations endow them with significant potential for temporal regulation. By flexibly adjusting charging schedules, these facilities can effectively participate in peak shaving and valley filling, thereby improving energy utilization efficiency. Fast-charging systems, in contrast, possess significantly higher power densities and rapid energy replenishment capabilities, offering unique advantages in terms of spatial deployment flexibility. They are particularly suited to emergency or urgent charging scenarios. However, current fast-charging technologies are constrained by thermal management limits and battery cycle life issues, and are thus primarily applied as supplementary solutions in emergency or special use cases [181].

4.2.3. Spatiotemporal Characteristics of EVs Load

Due to the spatiotemporal randomness of vehicle travel behavior and parking patterns, EV loads vary significantly across building zones with different functional attributes. As shown in

Table 4. Classification of EV Charging Methods [180].

	Standard Plug-In Charging	Slow Charging Station	Fast Charging Station
Current Type	AC	AC	Three-phase AC	DC
Power Rating	<3.7 kW	3.7–22 kW	22 kW−43.5 kW	<40 kW
Typical Locations	Residential homes	Offices, residential communities	Shopping malls, public venues

, this is reflected in the timing of peak loads, the structure of charging mode usage, and baseline initial State of Charge (SOC₀). EV charging power demand across different building zones exhibits three key characteristics: 1. Overall daytime load intensity is significantly higher than at night. 2. Daytime load is primarily from the office building charging infrastructure and public charging networks, while nighttime load is concentrated in residential zones. 3. Power demand at office charging facilities peaks during morning commute hours, whereas public service charging terminals peak during midday commercial activity [182].

Table 5. Spatiotemporal Characteristics of EVs Charging Demand.

Location	Peak Time [183]	Charging Type	Avg. Initial SOC Level [182]
Residential	Evening (6:00 PM–8:00 PM)	Plug-in/Slow-Charging Station	40.6%
Workplace	Morning (6:00 AM–10:00 AM)	Slow-Charging Station	47.8%
Public Venue	Midday (12:00 PM–2:00 PM)	Fast-Charging Station	39.1%

summarizes these spatiotemporal characteristics of EV charging demand. These variations stem from the deep coupling between building functions and user behavior. For instance, commuting results in vehicle concentration and morning charging peaks at workplaces; commercial activity drives midday public charging demand; and continuous nighttime parking in residential areas supports stable household charging loads. By aligning building energy curves with vehicle parking patterns, the spatiotemporal coupling of charging infrastructure can be enhanced, improving load aggregation potential on the building side.

4.2.4. Grid Regulation Services of EVs

EVs can support a wide range of grid regulation services, primarily including renewable energy integration, peak shaving, frequency regulation, and reactive power support: (1) Dynamic Power Regulation and Renewable Energy Integration. EVs, due to their rapid power response capabilities, can effectively smooth renewable energy output fluctuations. V2G technology coordinates distributed EV clusters with distributed energy systems, significantly enhancing the grid’s capacity to absorb high-penetration PV and wind power. Luo, L. et al. [184] demonstrated through a co-optimization model that V2G can increase the grid’s PV hosting capacity. Gao, S. et al. [185] proposed a renewable and adaptive integrated energy management scheme, where EV clusters dynamically compensate for wind power fluctuations to stabilize frequency and voltage. A spatial matching mechanism (prioritizing EVs near wind farms) further reduces regulation costs. (2) Load-Shifting and Peak-Shaving Services. EVs, with high grid connection rates, have strong spatiotemporal load-shifting capabilities. Bertolini, A. et al. [186] developed a deep reinforcement learning-based EV fleet charging scheduling strategy that, without compromising mobility needs, reduced peak-period EV load by 80%, mitigating the burden of large-scale EV integration. Xie, T. et al. [187] proposed a two-level optimization framework for “distribution network–charging station” systems, enabling coordinated demand response to maximize load modulation potential and validate the peak-shaving value of EV clusters. (3) Fast Frequency Response and Regulation Services. EVs’ millisecond-level power regulation capacity makes them ideal for frequency control. Alfaverh, F. et al. [188] proposed a deep reinforcement learning-based V2G control strategy that adjusts power dynamically while meeting user needs, enhancing frequency stability. Ma, S. et al. [189] proposed a model to optimize aggregator performance by balancing user satisfaction and frequency regulation revenues, highlighting how market mechanisms incentivize EV participation. (4) Reactive Power Support and Voltage Stabilization. Reactive power is essential for maintaining voltage stability [8]. Large-scale EV integration may cause node voltage drops, threatening grid safety [190]. By controlling charging modes and adjusting the real/reactive power ratios, EVs can provide voltage support services. Wang, J. et al. [191] developed a hierarchical coordination framework to optimize real power charging and reactive compensation strategies across multiple regions. Their results show this reduces user charging costs and mitigates voltage violation risks in distribution networks.

4.2.5. Influencing Factors of EVs’ Flexibility Potential

The flexibility potential of EVs is influenced by multiple factors, primarily user charging behavior, SOC levels, battery characteristics, and user psychology. Figure 7 illustrates the relationships among these influencing factors.

User Charging Behavior: As a subjective decision-making process, its spatiotemporal characteristics are jointly constrained by travel patterns and facility conditions. The selection of charging/discharging time slots typically follows two patterns: the “demand-driven” mode (charging when driving needs cannot be met) and the “anxiety-driven” mode (charging immediately after travel ends), leading to a characteristic dual-peak load curve in the morning and evening [192]. The choice of charging method (fast/slow) directly determines the intensity and duration of power impact on the grid. Areas with concentrated fast-charging infrastructure are prone to local peak loads. Charging frequency and geographical distribution shape spatiotemporal heterogeneity of load via spatial aggregation effects—the frequent use of fast-charging stations in commercial areas results in daytime load peaks, while nighttime slow charging in residential zones increases baseload pressure.

State of Charge (SOC): SOC is defined as the percentage ratio of remaining battery capacity to its rated capacity, and its dynamic evolution directly determines the power regulation capability and dynamic response characteristics of EVs in V2G participation. SOC management consists of three core control nodes: the initial SOC represents the vehicle’s energy status upon connection to the grid and directly affects the temporal distribution of load; the expected SOC reflects the user-defined target energy level, strictly constraining the depth of discharge to ensure travel requirements; the real-time SOC dynamically indicates the currently dispatchable capacity and adjusts the charging/discharging power threshold (±100% of the rated value) through real-time feedback mechanisms [193].

Battery Characteristics: Battery characteristics primarily concern capacity and lifespan. Battery capacity is positively correlated with regulation potential, with its rated value directly determining the upper limit of power regulation. Meanwhile, degradation in cycle life (manifested as capacity fade and increased internal resistance) constitutes a key constraint. Repeated charge–discharge cycles lead to structural degradation of electrode materials, resulting in capacity loss and increased internal resistance [194], which together impose dual constraints: capacity degradation weakens per-cycle charge/discharge capability and limits power regulation thresholds; lifespan degradation affects user behavior via economic pathways, significantly reducing willingness to participate in V2G [195]. The dynamic evolution of battery health (State of Health, SOH) thus becomes the core contradiction in balancing grid regulation demands and user economic expectations. It is necessary to optimize the battery operating range via intelligent charging/discharging strategies to mitigate the negative impacts of performance degradation on system flexibility.

User Psychology: User psychology mainly involves two core dimensions: strategic (game-theoretic) behavior and range anxiety. Strategic psychology refers to the behavioral game between electricity consumers and charging service providers under incentives such as dynamic pricing, which indirectly influences users’ charging/discharging decisions through demand response strategies. Range anxiety stems from users’ psychological expectations regarding insufficient energy reserves in the traction battery [196]. This psychological effect indirectly affects key parameters such as charging frequency [197] and the threshold setting of initial SOC (SOC₀) [198], and ultimately transmits to the spatiotemporal distribution of EVs charging load [199]. These two psychological factors jointly constitute the underlying logic of user charging behavior: strategic psychology dominates economically rational choices, while range anxiety reflects technical safety concerns. Their interaction shapes the typical behavioral patterns of EV charging loads.

4.3. Summary

As the two major flexible resources in buildings, air conditioning loads and EVs differ significantly in characteristics, regulation mechanisms, and application scenarios. Air conditioning load is a fixed thermodynamic regulation resource, using strategies like temperature settings, pre-cooling/-heating, airflow control, and chiller management to shift load over minutes to hours. Its core challenge lies in balancing comfort, thermal inertia, and rebound effects, requiring hierarchical control structures to optimize multi-timescale responses. In contrast, EVs are mobile bidirectional energy units that provide second-to-hour grid services (e.g., frequency regulation, peak shaving, reactive support) via V2H/V2G. Their regulation is constrained by user behavior (range anxiety or convenience), battery degradation, and SOC management, necessitating trade-offs between economic efficiency and battery health.

5. Typical Features of Flexible Resource Synergies

The preceding sections have provided a detailed analysis of individual flexibility resources. However, in practical applications, these resources, when considered individually, are often characterized by significant uncertainty, stochasticity, and limited capacity. These attributes severely constrain their potential to participate effectively in grid scheduling and interaction when operating in isolation [200].

To overcome these inherent limitations, it is imperative to aggregate and coordinately optimize these disparate, distributed flexibility resources from a systemic perspective. This approach aims to establish a large-scale and more dependable response capability for the grid [201]. Accordingly, this chapter focuses on the synergistic characteristics and application models that emerge from such coordination. We will use the energy community as a key analytical framework, conducting an in-depth exploration of its internal resource collaboration mechanisms to inform strategies for the coordinated dispatch of regional energy systems.

5.1. Energy Communities

Energy communities serve as a pivotal organizational framework and physical carrier for the coordination and aggregation of local flexibility resources. They are generally defined as collective entities voluntarily formed by energy consumers and prosumers within a defined geographical area to engage in joint energy activities [202]. Through internal coordination mechanisms, an energy community can operate as a single, unified entity, performing centralized optimal management and market interaction for its diverse assets, including distributed generation [203], energy storage [204], and controllable loads [205].

The primary advantage of this model lies in its capacity to effectively integrate and harness the complementary characteristics of its internal resources, thereby yielding direct economic benefits for its members [206]. To deeply analyze the underlying mechanisms that enable these benefits, this section will examine three core synergistic features: (1) the coordinated dispatch among multiple types of flexible loads within the community; (2) the source–load coordination between flexible demand and renewable energy generation; and (3) the enhancement of overall community operation through the integration of dedicated energy storage systems.

5.1.1. Coordinated Scheduling Among Various Internal Flexible Loads

Within an energy community, flexible building loads and EVs constitute the two principal types of controllable energy loads. However, when utilized as standalone demand response resources, each faces significant technical bottlenecks, making their coordinated dispatch essential for maximizing their aggregated value. Building flexible loads can be categorized into power-adjustable and time-adjustable types. Power-adjustable loads, such as HVAC systems, can shift energy use to assist in peak shaving, but their regulation depth is constrained by thresholds of human thermal comfort [207,208]. Time-adjustable loads, exemplified by appliances like washing machines, have time-shifting capabilities, but are limited by user-defined scheduling windows and small individual power ratings, placing a ceiling on their aggregate flexibility potential [209,210]. Both types face a common limitation—insufficient energy storage capacity, which restricts their ability to shift energy across time. Although EVs function as mobile storage units with spatiotemporal energy transfer capabilities, a mismatch exists between their charging demand and the timing of grid regulation needs. For instance, there is a time lag between the morning charging peak in office areas and the grid’s midday demand for load regulation [183], making it difficult for either resource to achieve global system optimization on its own.

Building–EV coordinated regulation overcomes these technical limitations through resource complementarity, enabling simultaneous improvement in energy economy and grid stability [211]. Figure 8 provides a schematic diagram of this synergy. Coordination between power-adjustable loads and EVs focuses on optimizing power parameters. Multi-objective decision models are developed to balance key performance indicators, such as charging costs, operational costs, grid loads, and user comfort [211,212]. A representative example is the joint operation of HVAC systems and EVs, which store energy (thermal and electrical) during off-peak periods and release it during peak hours [213,214]. This cross-medium energy coordination significantly enhances regional energy efficiency. Furthermore, regulation strategies must incorporate device prioritization mechanisms to guarantee energy provision for high-sensitivity equipment and maintain user satisfaction. The synergy between time-adjustable loads and EVs focuses more on time sequence optimization. The time-shifting flexibility of building loads helps fill control gaps when EVs are offline. When load flexibility is saturated, controlled EVs charging and discharging are used to maximize marginal benefit. For instance, Cetinbas, I. et al. [215] coordinated the operation of EVs and shiftable loads, demonstrating that co-optimization achieved approximately 36% in monthly electricity bill savings compared to EV-only scheduling. Khemakhem, S. et al. [216] proposed a two-layer energy management strategy that prioritizes building flexible loads for peak shaving and valley filling, with EVs providing supplementary flexibility through their charging and discharging capabilities, thus achieving both load balancing and power system stability.

5.1.2. Synergy Between Flexible Loads and Renewable Energy Sources

The successful, widespread deployment of distributed renewable energy hinges critically on public acceptance and active participation. By enhancing the capacity for local self-consumption and optimizing energy efficiency, energy communities directly reduce energy costs for their members, thus serving as a pivotal mechanism for promoting the deployment of distributed renewables [202]. Consequently, establishing a multi-dimensional coordination framework that integrates flexible building loads, EVs and photovoltaic (PV) systems is a critical pathway to enabling the efficient, localized consumption of renewable energy within the community. This synergistic framework is depicted in Figure 9. For example, Srithapon, C. et al. [217] proposed a coordinated scheduling model combining heat pump-based thermal storage and EV charging/discharging, which significantly increased PV self-consumption in summer, with a 32.97% increase in PV self-use. Golshannavaz, S. et al. [218] introduced a peak-shifting strategy involving time-adjustable loads and EV charging. Through load sequencing optimization, they enabled efficient utilization of PV generation and suppressed grid peak loads. Harder, N. et al. [219] found that in residential contexts, integrated operation of EVs, energy storage systems, thermal storage, and HVAC equipment can significantly smooth out building load profiles.

Furthermore, the inherent mobility of electric vehicles (EVs) introduces a unique spatial dimension to the synergies within energy communities, enabling them to act as mobile energy couriers between different locations—a concept often termed ‘Vehicle-to-Building-to-Building’ (V2B2). Alruwaili, M. et al. [220] detailed the spatiotemporal characteristics of EVs and e-bikes, incorporating them as mobile, bidirectional storage units into the coordinated scheduling of an energy community. This approach achieves efficient temporal energy shifting, enhancing regional energy flexibility while reducing operational costs. Leveraging this mobility for spatial flexibility, Barone, G. et al. [177] demonstrated that EVs can acquire and store surplus energy at one location (e.g., a PV-equipped office building during the day) and, through the user’s daily commute, physically transfer it to another location with an energy deficit (e.g., a residential building in the evening). This strategy transforms isolated building clusters into cooperative energy communities, breaking the energy self-balancing constraints of individual buildings and significantly increasing the shared utilization and consumption of regional renewable energy.

These findings collectively highlight the core value of multi-energy flow synergy—by leveraging the spatiotemporal complementarity of EV-based mobile storage, building-side load flexibility, and PV generation, a dynamic “source–load–storage” balancing mechanism can be constructed. This mechanism not only increases PV self-consumption in buildings but also enhances overall system flexibility through cross-medium energy conversion and multi-timescale coordination, offering innovative solutions for developing high-resilience smart energy systems.

5.1.3. Communities Equipped with Dedicated Energy Storage Systems

The aforementioned strategies for internal flexible load coordination and source-load synergy form the foundation of an energy community’s flexibility. However, due to the high uncertainty inherent in both flexible loads and distributed renewable energy sources, communities relying solely on these resources struggle to provide high-value ancillary services to the grid. Therefore, integrating dedicated energy storage systems (ESSs) within energy communities is a common strategy to overcome these bottlenecks. For instance, Fotopoulou et al. [206] treated an ESS as a shared asset among community members. Through centralized optimization of the ESS charging and discharging schedule in a day-ahead context, they effectively reduced reliance on the external grid and minimized electricity procurement costs, thereby minimizing the community’s overall operational expenditure. Manso-Burgos et al. [221] positioned energy storage as a key component within a portfolio of flexibility resources in a highly electrified energy community. Their study demonstrated that the coordinated dispatch of PV, storage, and various flexible loads significantly enhances local energy self-consumption, thereby maximizing the community’s economic benefits. They also noted that the profitability of battery storage and PV systems is heavily dependent on the community’s level of electrification. Furthermore, M. Sani et al. [222] incorporated a hydrogen energy system as a core component to complement intermittent renewables like wind power, thereby improving the reliability of the community’s energy supply. Their research revealed the significant potential of hydrogen for long-duration energy shifting, which effectively boosts the community’s energy self-sufficiency and supply reliability, and may even reduce dependence on the main grid and lower operational costs in the future.

In summary, while dedicated energy storage is crucial for enhancing community flexibility and economic viability, current research and practice exhibit an over-reliance on costly, single-vector electrical storage. This reliance constitutes a significant barrier to widespread deployment. Therefore, future research must pivot towards diversified, multi-vector energy storage strategies. This necessitates not only integrated planning at the technical level—to determine the optimal capacity, power rating, and technological mix of electrical, thermal, and chemical storage for a balanced cost-performance profile—but also bold innovation in business models. New mechanisms must be explored to stack multiple value streams and effectively amortize high initial capital costs, thereby enabling the development of economical, efficient, and truly sustainable energy communities.

5.2. Field Research on Energy Communities

The theoretical and model-based analysis in Section 5.1 systematically established the fundamental principles of energy communities, illustrating their immense potential for optimizing local energy resources and realizing economic benefits through internal synergies. To translate this theoretical promise into reality, an increasing number of field studies and commercial deployments worldwide are providing critical empirical evidence.

For instance, a prominent field demonstration is the Brooklyn Microgrid (BMG), a project initiated by LO3 Energy in New York City aimed at addressing the grid reliability challenges posed by the high penetration of renewable energy and reducing dependence on the main grid [223]. The project introduced a peer-to-peer (P2P) energy trading mechanism, thereby establishing an internal energy market within the community. This market not only enabled the synergistic optimization of conventional user loads but also successfully integrated emerging flexible loads, such as EVs charging stations, into the coordination framework, thus effectively promoting the local generation, storage, trading, and consumption of renewable energy. The operational results proved that this internal community coordination significantly enhances the resilience and reliability of the regional grid, providing a replicable and successful model for the development of future urban energy systems.

Similarly, the Storage ENabled SustaInable Energy for BuiLdings and communitiEs (SENSIBLE) project, conducted jointly in Portugal, the UK, and Germany from January 2015 to December 2018 [224], improved local capabilities by integrating diverse storage technologies into local energy networks, homes, and buildings, and by linking these assets to energy markets. This approach led to effective enhancements in the local consumption of renewables, regional power quality, energy flexibility, and economic benefits. It provided a key technical blueprint and invaluable operational experience for developing future energy communities characterized by highly interactive demand-side resources [225].

In summary, these pioneering practices collectively point to a core conclusion: establishing energy communities to enable the effective coordination of regional flexibility resources and the efficient local consumption of renewables is a key pathway for transitioning future energy systems towards a clean, efficient, and economical paradigm.

5.3. Summary

The coordinated operation of distributed renewable energy, EVs, and building flexible loads enhances system flexibility and renewable energy utilization efficiency through multidimensional complementarity. In the DRE–EVs synergy, DRE offers low-cost electricity for both buildings and EVs, while the energy storage capabilities of EVs allow for intertemporal energy transfer to mitigate curtailment of wind and solar resources. However, the dynamic matching of intermittent DRE output and uncertain EVs charging behavior remains a challenge, requiring real-time control frameworks integrating stochastic optimization and predictive scheduling. EV–building flexible load collaboration overcomes the limitations of individual resource types through complementarity: Power-adjustable loads (e.g., HVAC) and EVs can jointly store and release energy to smooth peak-valley imbalances while maintaining user comfort. Time-adjustable loads can stagger their operation with EVs charging/discharging to fill control gaps and reduce electricity costs—co-optimization has been shown to save up to 36% in monthly electricity expenses. Further integration of DRE, EVs, and building-side flexible loads (e.g., EVs + thermal storage + PV) supports the development of a dynamic “source–load–storage” balancing mechanism, enabling a 32.97% increase in PV self-use and improved load profile smoothness. The core value lies in utilizing spatiotemporal complementarity across mobile energy storage, flexible demand, and multi-energy flow conversion to overcome technical boundaries, enhance system resilience, and improve energy economics—thus providing innovative pathways for smart energy systems.

6. Regulatory Frameworks and Case Study

The effective mobilization of the flexible resources analyzed in previous chapters depends critically on robust regulatory frameworks and economic incentives. Chief among these is Demand Response (DR), which facilitates the modulation of end-users’ inherent electricity consumption patterns in response to dynamic signals, such as electricity prices or explicit dispatch requests, to support grid stability and efficiency [196].

Addressing the need for a practical and regulatory perspective, this chapter provides a structured overview of the principal DR mechanisms currently in practice. We begin by presenting a comprehensive taxonomy in

Table 6. Classification and Comparison of Demand Response Schemes.

Type	Definition	Main Characteristic	Main Shortcomings	Typical Application Scenarios	References
Price-Based Demand Response
TOU	Electricity prices vary across pre-defined time blocks (e.g., peak, off-peak) to encourage users to shift consumption to off-peak periods.	Simple to understand and easy to implement.	Can create new load peaks during low-price periods; Poor responsiveness to dynamic grid fluctuations.	Applicable to users of all scales, including large commercial and industrial (C&I) and residential customers.	[226,227]
RTP	Prices are adjusted day-ahead or intra-day based on market supply-demand fluctuations to guide user consumption patterns.	Provides dynamic incentives for load shaping that reflect real-time grid needs.	Relies on real-time, two-way communication; effectiveness is limited by data exchange and smart metering infrastructure.	Mainly suitable for industrial and commercial sectors; less common in the residential sector.	[228,229]
CPP	A significant surcharge is added to the standard price during a few critical grid stress events to incentivize drastic load reduction.	Enhances power system reliability by preventing load from exceeding grid capacity during emergencies.	CPP events are dispatched only during extreme grid stress, which may occur only a few times per year.	Mainly suitable for industrial and commercial sectors; less common in the residential sector.	[226,230]
Incentive-Based Demand Response
DLC	Based on a pre-agreed contract, a utility or third-party aggregator can remotely control end-use appliances.	Enables rapid dispatch with high reliability.	Centralized control poses significant cybersecurity risks and raises user privacy concerns.	Suitable for residential and small commercial customers (e.g., HVAC, water heaters).	[231,232]
DB	Allows users to participate in electricity markets by submitting bids with prices and load curtailment volumes to provide flexibility services.	Empowers users with decision-making autonomy and control over their equipment, leading to higher engagement.	Involves complex market mechanisms and high uncertainty in participant behavior.	Primarily used in commercial and industrial sectors; less common in residential.	[233,234]
I/C	During system emergencies, the operator can curtail the power supply to consenting customers according to a pre-signed contract.	Provides highly reliable, contractually guaranteed emergency reserve capacity.	Can be economically inefficient and may incur high costs due to production losses for the customer.	Suitable for large power consumers (e.g., >200 kW).	[235,236]

, which systematically categorizes existing DR schemes—distinguishing between price-based and incentive-based approaches—and compares their core characteristics, operational limitations, and typical application scenarios. Building upon this comparative framework, the subsequent sections delve into these strategies with a focus on real-world implementation. To illustrate their practical efficacy and impact, we will examine prominent case studies from various international contexts.

6.1. Price-Based Demand Response

Price-based demand response operates by conveying time-differentiated market signals to energy users, leveraging economic incentives to motivate changes in their consumption patterns. Many countries and regions have widely implemented time-of-use (TOU) tariffs, which differentiate electricity prices by hour, day, or season. These tariffs encourage end-users to reduce electricity consumption during peak demand periods, thereby achieving grid objectives like peak shaving and valley filling. For instance, national policies in China mandate the establishment of a comprehensive TOU framework—encompassing peak-valley, sharp-peak, and seasonal rates—to accommodate the large-scale integration of renewables and ensure power system security [226]. Similarly, France employs the “Tempo” tariff structure, which features six distinct price levels organized across three day-types—blue (approx. 300 days/year), white (approx. 43 days/year), and red (approx. 22 days/year)—each with its own peak and off-peak periods to influence user consumption [227].

However, the static, pre-defined time blocks of traditional TOU tariffs are often inadequate for addressing the volatility challenges introduced by high penetrations of renewable energy. Consequently, many jurisdictions are exploring more dynamic mechanisms like real-time pricing (RTP). The first RTP programs were implemented in the southeastern and midwestern United States in the 1990s, announcing prices on a day-ahead or hour-ahead basis to prompt users to adjust their energy use [228]. According to data from the U.S. FERC and EIA, there are currently 44 RTP programs offered by 29 utilities across 24 states, including by major providers like West Penn Power Company and Commonwealth Edison [229].

Numerous studies and practical implementations have demonstrated that PBDR strategies effectively engage demand-side flexibility to achieve grid regulation goals. A study in China found that in 2014, TOU tariffs shifted 21.86 GW of demand, covering 73% of the national power deficit and increasing the grid’s average load factor to 85.72% (a 1.84 percentage point improvement over 2013) [237]. It is estimated that France’s “Tempo” tariff reduces consumption for an average 1 kW household by 15% on “White” days and by 45% on “Red” days [238]. In the United States, the 15,436 residential, commercial, and industrial customers participating in RTP programs have achieved a potential aggregate peak reduction of 1905 MW [229].

6.2. Incentive-Based Demand Response

Incentive-based demand response elicits changes in energy consumption through explicit agreements, established via contracts or bidding processes, between a provider (the consumer or a designated third party) and a buyer (a market participant or grid operator). In contrast to price-based approaches, this model circumvents the reliance on consumers’ purely voluntary responses to price signals, thereby offering a more direct and dispatchable form of flexibility [239].

IBDR programs are predominantly classified into two main categories: Interruptible/curtailable (I/C) services and direct load control (DLC). I/C is a DR modality wherein, during periods of grid stress or when load regulation is needed, the system operator can request that consenting consumers temporarily curtail their power supply or reduce consumption according to a pre-agreed contract [235]. Several countries and regions have adopted such strategies. For instance, Finland’s national Transmission System Operator signs contracts with industrial users for an annual interruptible program that serves as a disturbance reserve; in 2005, this DR potential was estimated at 1280 MW, representing 9% of Finland’s peak demand [236]. In Germany, the Ordinance on Interruptible Load Agreements (AbLaV), established in 2012, permits interruptible loads to participate in the balancing market as secondary and tertiary reserves. As of 2020, 20 such contracts had been signed, covering 2.5 GW, of which 0.8 GW was classified as interruptible load [240]. DLC involves an agreement where a user authorizes a central controller (such as a utility or an aggregator) to remotely manage the operation of their equipment in response to a DR signal, in exchange for a direct incentive [231]. A pilot project implemented in Sala, Sweden, from 2011 to 2013 demonstrated and evaluated the potential of DLC strategies to improve local grid efficiency [232]. In the United States, DLC has become a mature commercial strategy, widely applied in the residential sector. Utilities offer incentives such as annual rebates ($25–$100) or performance-based payments (per kWh saved). The most common application involves the automated, short-duration cycling of appliances like air conditioners via smart thermostats or dedicated switches during peak hours, enabling rapid and precise system-wide load reduction [241].

7. Conclusions

In the context of low-carbon energy transitions and high penetration of renewables in power systems, this paper systematically reviewed flexible resources in building energy systems and interactive technologies between EVs and the grid.

For regional energy systems, it focused on analyzing the flexibility characteristics, core challenges, and solution paths of distributed energy supply technologies (distributed renewable energy, building-integrated cogeneration), distributed energy storage technologies (electrical and thermal/cooling storage), and building flexible loads (air conditioning loads and EV loads). It also summarized the collaborative features among typical flexible resources.

The main conclusions are as follows:

• Distributed renewable energy and building-integrated cogeneration technologies form a complementary flexible energy supply system. The former smooths intermittent output fluctuations through storage systems, while the latter builds electricity–heat–cooling cogeneration based on energy cascading, jointly forming a “renewables–multi-energy cogeneration–storage” integrated architecture that enhances building energy efficiency while meeting multi-scale grid regulation demands.
• Distributed storage achieves complementary advantages through the differentiated synergy of electrical and thermal energy storage: fast-response electric storage supports instantaneous power, while cost-effective thermal storage enables temporal energy shifting. Their integrated optimization is central to constructing a flexible regulation system.
• Among building flexible resources, air conditioning loads achieve hour-scale regulation via thermodynamic control, requiring a balance of comfort and thermal inertia. EVs, as mobile energy storage units, participate in grid services through V2X modes but are constrained by user behavior and battery degradation. Their synergy breaks individual resource limits: power-adjustable loads coordinate with EVs storage to balance load, and time-adjustable loads stagger with EVs charging/discharging to reduce costs.
• Integrating DRE, EVs, and flexible loads into a “source–load–storage” dynamic balance mechanism significantly improves system flexibility and economy via spatiotemporal complementarity, providing innovative solutions for smart energy systems.
• To fully exploit building energy flexibility in support of high renewable energy integration, future work needs to further develop a quantitative characterization and scale expansion of the “spatiotemporal complementarity” framework. This includes establishing a mathematical indicator system that accurately describes the load complementarity among building clusters, and exploring the upward extension of this framework to enable coordinated optimization with transmission and distribution network planning and operation, thereby enhancing the overall efficiency and flexibility of the energy system.