1 Introduction

1.1 Background

Energy is not only a vital support of modern economy but also the basis for the survival and development of human society. Since industrialization, fossil fuel such as coal and oil has become an indispensable energy basis for human society. However, with the increasing of energy consumption, the shortage of traditional energy has caused serious problems. In the Stated Policies Scenario [1], the growth speed of electricity use is more than a double of overall energy demand, confirming its critical role in modern economies. The share of electricity which is less than the half of oil today, will overtake oil by 2040 [1].

In order to alleviate the situation of energy shortage and the adverse impact of traditional energy on environment, many scientists adhere to the research and development of renewable energy. Progress in the power sector was concentrated while renewable energy is gradually becoming more economical compared to the conventional thermal power. By the end of 2019, renewables provided an approximate more than 26% of global electricity generation [2]. Among them, power generation technologies of renewable resources e.g., solar energy and wind energy, have attracted wide attentions and become the main sources of electricity supply. In the mid-2020s, the advancement of wind and solar PV generation outstrips coal in the power generation mix [1]. More than half of the total electricity generation will be supplied by low-carbon sources by 2040, in which wind and solar PV will play the key roles [1].

However, these renewable resources are volatile and unpredictable for power generation. Even with a well monitored and controlled smart gird, the high variability of renewable energy resources requires platitudinous energy storage [3]. With the technological development and cost reduction of electrical energy storage (EES) recently, EES could be installed to optimize the performance and stability of renewable systems [4]. The development of battery energy storage system (BESS) technology is found to be critical to the system volatility and unpredictability [5]. In addition, BESS can ameliorate the efficiency of energy utilization [6] and decrease the impact of peak demand period caused by the traditional power grid [7]. Many countries have put forward different electricity pricing policy scenarios to encourage users to use BESS [8]. From the view point of users, it is the most attractive to use batteries to reduce daily expense. Because the price of batteries is going down and their performances are improved [9], distributed electrical storage shows great potentials not only in independent units but also in the electric vehicle fleet [3].

As a top energy user, building energy consumption should also be further optimized. With the continuous improvement of built environment, building energy consumption is increasing compared with the situation decades ago [10]. It is of great significance to use BESS in buildings and even to connect it with other networks to form a more effective system for energy utilization [11] and achieve building energy flexibility (BEF). Besides, the combination of photovoltaics, building energy saving and battery storage systems has also proven to be a critical means to mitigate the catastrophic climate change risks [12]. The BESS will certainly occupy a more prominent position in the future for its use in buildings [13] and make bigger contributions for BEF.

1.2 Motivations

It is a critical question that how BESS can be better built and managed, not only for meeting the building energy demand but also for increasing the building energy flexibility (BEF). Energy consumptions of different buildings will vary greatly with time and building function [11]. BEF is a new concept for solving this problem. Compared with the rigid building load, the flexible building load is obviously more suitable for the actual user’s demand. Considering the noticeable importance of BEF and BESS, this review study is motivated by the reasons summarized in Table 1.

Table 1 Analysis on research gaps
Full size table

Concerning the study on BEF with BESS, the design, choice, and control of battery [14, 21] is merely a small part of the whole project. In this study, our motivation is not limited by offering new advances in relevant topics, but also providing new insights and outlooks for the future development by identifying the key difficulties. Through this study, the proposed questions, suggestions, and outlooks may trigger more and more researches which could boost the healthy and fast development in building sector and beyond.

1.3 Organization

The remainder of this paper is organized by a logic chain as: Sect. 2 first presents some basic and general information and status of BEF which may help readers grasp the big picture. Section 3 provides the features and requirements of BEF from the angle of demand side. Based on those features, Sect. 4 discussed the system of power supply and battery from the perspective of renewable energy power generation and the power consumption system covering mobile batteries (electric vehicles). In addition, Sect. 5 gives some discussions on critical concerns and outlooks for BESS and BEF, and Sect. 6 has a final summary of this study by noting the major contributions, focuses, targeted issues as well as some suggestions for future developments.

2 Overview of building energy flexibility (BEF)

2.1 Basics

Building energy flexibility (BEF) has not been precisely defined yet. In general, BEF refers to the load with flexible characteristics that can actively participate in power grid operation control and interact with power grid [22]. The concept of flexibility means the capability to preserve balance over energy generation and load (i.e., energy consumption) under uncertainty of the power system [23]. The BEF is variable within a certain period, in which the load uncertainty [24] should be seriously considered since it has noticeable impact on the results. The building load could be flexible through a resource regulation and dynamic power management tool [25]. And in order to achieve flexible electricity loads, energy storage systems play an important role for load shifting or adjustment [26].

Depending on the forms of energy usage, both thermal energy storage [27,28,29] and battery-based electricity storage are critical for BESS as shown in Fig. 1 [30]. For thermal energy storage, it is a novel idea to use building structure and furniture for heat storage [31]. As for battery-based electricity storage, the regulating effect of battery storage on building energy consumption [15] and the regulating ability of battery storage on power grid [7] all show significant impacts.

Fig. 1
figure 1

Illustration of different flexibility and energy storage options within a building [30]

Full size image

In the overall building energy system, the demand-side response (DR) is closely linked with BEF and BESS. The development of battery promotes the DR, and it can also optimize the use of battery in return [32]. DR offers a means of settlement to the grid imbalance which place restrictions on the use of renewable energy. DR could contribute to the stability of smart grid while succeed in cost savings for buildings and other participants in the smart grid [25]. At present, there are many effective models for DR according to different types of buildings such as commercial buildings [33], office buildings [34], residential buildings [35].

2.2 Research trend and distribution

Since the BEF is an increasing popular topic in the building and energy field, the development trend and distribution need to be identified. The statistical data of the publication years and the nationality of authors are obtained as shows in Fig. 2 by using the database in the Web of Science with the keyword “building energy flexibility”. The statistical results show that the number of publications of BEF is increasing year by year. Especially in the past decade since 2010, the ascending speed is becoming faster. In fact, from the perspective of literature content, the relevant studies have also been further expanded. The momentum behind this increasing research interest is related to severe energy problems, renaissance of artificial intelligent, evolvement of smart grid as well as some impacts from government policies.

Fig. 2
figure 2

Yearly publications related to BEF and country distributions of authors (data source: Web of Science)

Full size image

It is noteworthy that the top six publishing countries account for nearly 90% of the total publication number for BEF. The percentage of China’ contribution is 41.8% which is far more than the second place USA. Ranking as the first place, China plays an important role and is making bigger contribution now. According to the publication data and the trend, it can be forecasted that the publication number in this topic could be over 700 in 2022.

3 Building energy demand and response

3.1 Building energy demand

A critical analysis on the load features of different building is the basis for a better demand response (DR) management towards BEF. Public buildings have the technical ability to provide a lot of flexibility for the electricity system [36]. The HVAC systems in public buildings can be used as a suitable demand response resource to provide auxiliary reserves for several reasons. The demand response of HVAC system is much faster than the 10 min warming-up fully responding for generators, and the curtailment can be achieved almost immediately [37]. Therefore, when considering energy conservation by demand response management in public buildings, HVAC system is often the main object [38].

While the flexible electricity demand in residential buildings suggests theoretically flexible loads of two categories. The first category includes the loads with evident thermal inertia, such as air conditioning and refrigeration, space, and water heating. The second category includes loads with separation of electricity and energy services demand in time. The so-called ‘wet’ implement: dishwashers, washing machines, and tumble dryers are included in this category. Consumers ought to change their demand for energy services such as lighting, cooking and entertainment, which will not be as flexible as they are involved in demand response [39].

The industrial building sector is particularly suitable to implement DR for BEF, by shifting industrial electricity consumption to night time based on the production need [40]. Further, most of the manufactured products are much easier to be stored. There is an important distinction between industrial “production” loads and “support services” load. Support serves loads can usually be used more flexibly than the production loads [41]. In the residential and commercial buildings, the control of HVAC and lighting systems can realize the load adjustment, whereas industrial demand side management (DSM) involves power-intensive industrial processes with controlled production levels [42].

3.2 Building energy demand response

3.2.1 General classifications and tools

Demand response (DR) [39] can be realized by a series of customers’ actions according to specific conditions in the system. End users may need to achieve temporary reducing, increasing, or shifting power consumption by changing the plan of energy consumption, or using on-site generation and energy storage rather than directly taking electricity from the grid [36]. DR is helpful to combine new decarbonization electric loads and different renewable sources [39]. Its arrangement helps to optimize the utilization of existing infrastructure, stabilize the price of electricity market, and decrease the future investments demand of grid and peak load.

DR can be classified based on the timing, frequency, duration, and its predictability and on whether response happens due to manual behavior changes, direct load control, or automation. They may also be sorted as static or dynamic state, depending on whether pricing or other signals follow a preconcert schedule [39]. Distinction between static and dynamic interventions means a lot. The dynamic interventions might change continuously instead of following a predetermined schedule [43].

Researchers propose different control models for DR in different types of buildings, and energy flexibility is gradually considered in actual use [17]. The DR model based on BEF can be also discussed in terms of economy [44]. Possible business models for energy efficiency and DR providers in various electricity market sections are analyzed. The analysis shows that the demand side surge of energy storage resources will ensure more business models working with the electricity market department. DSM providers can also choose appropriate business models according to the electricity market and the DSM duration/interruption frequency tolerance, ramp ability and required advance notice time [44]. Diverse DR programs may lead to different risks, rewards, and uncertainties. Finding a balance between them may lead to effective DR control solutions [25].

As shown in Fig. 3, under the scheme of dynamic DR, it can be further divided into two broad categories. In the first type of “incentive-based DR programs”, consumers get incentives to change their energy consumption patterns based on the supply-side requirement. In the second one of “price-based DR programs”, consumers are charged differently at different time. Thus, the retail electricity tariff flows with the supply cost [45]. In general, complex requirements lead to comprehensive tariffs. People usually choose one control mode or couple it with other modes to achieve the optimal effect under the coincident circumstance.

Fig. 3
figure 3

Classifications of demand reduction and demand response [43]

Full size image

Table 2 shows the classifications and the description of common and specific types of pricing and other economic incentives excerpted from the Ref. [43].

Table 2 Types of pricing and other economic incentives in the Ref. [43]
Full size table

3.2.2 DR with building energy storage

When the building energy flexibility (BEF) is concerned with a better control and balance between the supply and demand sides, many complex factors should be considered. Ottesen and Tomasgard [46] proposed a model to support building energy flexibility through one energy hub for the scheduling of multiple strategies. The Ref [17] summarizes several improved strategies for BEF, which are classified and shown in Fig. 4.

Fig. 4
figure 4

Classification of improved strategies for electricity flexibility [17]

Full size image

In public buildings, DR is commonly considered to decrease consumption at a specific hour or season. Thereby the grid peak demand is decreased. In the field of residential buildings, complex appliance-level controls such as on-demand reduction of air-conditioning or automatic dimming of lights are less common [47]. Technologically, energy storage and bi-directional-inverters are conveniently available for household DR [47].

In fact, utilization of the energy storage system is inevitable when DR is considered in a building. As a flexible energy resource commonly referred, battery storage can be discharged at peak load and charged at load-valley time [48]. Generally, as one of the concerns from owners, the battery has a relatively short life span and needs high investment [49]. However, in BESS, the system integrated with EV, battery storage and roof-top PV, has gradually produced profound changes. Correspondingly, the declining price of PV system and battery will make battery storage system more affordable and appealing [17].

4 Renewable power generation and battery storage for BEF

Renewable energy systems (RES) are restricted to access to the main grid due to inherent shortcomings such as the fluctuation and intermittence of output power, inconsistency with load curve and the influence on the relay protection [50]. In order to alleviate those problems and achieve a better energy-saving effect, building battery storage is usually coupled with renewable energy generation. Common modes are the PV-battery system, the wind-battery system and the PV-wind-battery systems which are integrated together.

4.1 PV-battery system

The combination of PV system and battery increases the utilization of solar radiation [51] and the on-site self-supply for buildings [52]. Hence surpluses diurnal power from PV is stored to provide the electricity at night-time [53]. The integration of BESS can help to ensure maximum PV self-consumption and decrease the burden of distributed renewable resource grid, which allow trade-in of extra PV power [54].

Nowadays, in the market, there are abundant PV-battery systems applied in residential buildings. A PV-battery system generally is composed of several PV modules, power inverters, a battery storage device, and a control component. The battery can be connected to the AC- or DC-side of the PV inverter [53]. The AC-coupled BESS adopts bi-directional battery inverters while the un-directional inverters are always combined with DC-coupled systems. Nevertheless, there are also DC-coupled systems with bi-directional inverters that can be charged from the AC-side [55].

Figure 5 displays a schematic diagram of a PV system in electricity distribution grid. In this instant, the system is only for one user while it can also be extended to a group of users. When considering the connection between PV systems and the power grid, some basic orientations of energy transfer are the basis for adding value to its application [54].

Fig. 5
figure 5

PV system linked to the grid [54]

Full size image

As the main part of a standalone building with PV system, the PV-battery system may make a great improvement in energy efficiency [56]. But the factors that need to be considered for PV-battery system control are very complex. Niu et al.[57] proposed an autoregressive model with exogenous inputs that predict the dynamic cooling demand and using a mixed integer linear model to improve the scheduling of building energy systems with minimum operating expense. It is shown that without damaging the power feeder stability, the operation cost can be also greatly declined.

In addition to PV-battery system, this system may be integrated with building thermal energy storage. The Ref. [58] shows the use of PV-battery is feasible and has potential advantages to improve the thermal performance of building envelope. Häring et al. [59] presented a study of a general thermal storage system that is used for DSM in off-grid cases for near zero energy buildings integrated with PV-battery system. The results show that the battery capacity can be reduced significantly when thermal storage is of concern.

4.2 Wind-battery system

For the wind-battery system, the electricity generated by wind varies with weather conditions. Compared to the PV system, the wind system has higher application requirements. On a preset day, the energy contribution of wind to the network can vary significantly. When the amount of contribution is low, the electricity from the network is supplied by traditional power production. When this contribution is high, a specific ratio of the energy generated by wind may be wasted because of curtailment [60]. Only in some areas with abundant wind power, wind power generation would be considered in micro-grid or individual buildings.

At present, the use of wind power in buildings is to build a large-scale wind plant in the area with abundant wind resource. Even with the incremental installation capacity of wind power, wind power generation is still uncertain and intermittent. In addition, the wind energy forecasting is very difficult, which imposes an enormous influence on the regional grid. However, the wind generation and other energy sources can relate to buildings by the means of the assistance of BESS as displayed in Fig. 6 [60].

Fig. 6
figure 6

Multi-objective optimization strategy with an integrated micro grid in a grid-connected building [60]

Full size image

BESS is a strong way to identify the problem of large-scale grid connection with wind power because BESS can make up for the issue of inaccurate forecasting of wind energy production. The joint operation of a wind power plant and BESS is helpful to track the energy generated by wind, thus optimizing the effectiveness of wind power usage. The combined system may also be beneficial to the grid stability when a large-scale wind power plant is linked [61]. Different DSMs (demand side management) are tried to optimize the wind-battery system towards BEF. Since price-based DSM urges load shifting instead of load shedding, the study [62] show that the simultaneous percentage share of wind is lessened, which defers or averts the demand on wind power by moving load to the time when wind energy is highly available. This study [62] further shows the DSM system may instantaneously decrease the requirement to fossil fuel generation, and the utilization influence on the devices. In the best case, the carbon emissions can be decreased by about 4.21%. In another study [63], a model for the energy management of a residential building network interacting with wind energy conversion and BESS was established. In this study, the Feed-in and Time of Use (TOU) tariff were used, and the results demonstrate that the decrease of cost is dependent on variables such as the size of the wind energy conversion system and the energy storage system, the load profile, as well as the availability of the resource.

Ai et al. [64] pointed out model predictive control -based (i.e., MPC-based) strategy was a good choice for facilitating the practical use in real-time system, and the system performance can be optimized by using dynamic prediction according to the SOC (state of charging) of BESS. The numerical results showed that the methodology is effective. Commonly, adding a BESS can boost the flexibility of wind farm and a symmetrical regulation reserve.

4.3 PV-wind-battery system

Renewable power generation is volatile. The obvious intermittence and uncertainty of RES may result in unreliability on building load demand by merely using one source. A hybrid PV-wind system (HPWS) may weaken the uncertainty problem, while the combination of a BESS (hybrid photovoltaic-wind-battery system HPWBS) alleviates the intermittence problem [65]. The hybrid system is more reliable and can even installed in remote areas. When the energy generation capacity is more or less than the required load, the difference can be varied with the public network through net metering service. In addition, hybrid systems in the region can applied without reliable power sources. Therefore, it can be an emergency system to replace the conventional uninterruptible power supply (UPS) in a long period.

The hybrid residential energy systems are based on PV panels, wind turbines, and/or micro-turbines gradually that attract attentions and receive applications. However, such a system should be combined with an energy storage solution, most usually a battery. Figure 7 presents the scenario of the grid-connected HPWBS for a residential building [66]. This plant comprises a PV generator with an DC/AC converter, a wind micro-generator with an AC/DC rectifier, an electric battery storage system, a regulator, and a DC/AC inverter. Solar PV and wind turbine can meet the power demand of the building based on design and initial calculation. When the load exceeds what the RES can offer, the battery bank will be discharged. In contrast, when renewable energy output is higher than demand, and the state of charge (SOC) of battery is over 95%, the residue will be converted into thermal energy. In such a model, thermal energy can also be further utilized [66]. Achieving the best SOC of battery as much as possible is the focus of battery utilization [67].

Fig. 7
figure 7

The energy flow diagram of a hybrid (PV-wind) system with battery and thermal storage [66]

Full size image

For the purpose of environmental protection by minimizing the life cycle cost (LCC), dump energy and CO2 emissions, Ogunjuyigbe et al. [68] proposes a tri-objective design of PV/Wind/Split-diesel/Battery hybrid energy grid independent system with genetic algorithm (GA) for a representative residential building. The results reveal that the PV/Wind/Split-diesel/Battery is the best solution, and its LCC is $11,273. Compared to the single diesel generator scheme, the LCC, COE, CO2 emission and dump energy offers 46%, 28%, 82% and 94% reduction respectively. This hybrid system in another study [69] decreases the CO2 emissions for all given locations by adding the battery capacity, and the decrease is more significant in the locations with higher solar potential. But at the same time, the higher battery capacity increases the levelized cost of electricity (LCOE) violently. A balance should be reached in the real application according to the local condition.

Taking the advantage of the flexible feature of a micro-grid system, the regularity and randomness of the renewable sources (solar, wind) can be modeled by the adaptive control and the stochastic method. Sarkar et al. [70] presented a distinctive integration of solar PV, wind, biomass, and vanadium redox flow battery (VRFB) storage, which is a hybrid micro-grid. The capacity selection and techno-commercial optimization of different RESs to meet the daily energy demand was studied. Moreover, the peak load regulation can be built by providing the actual data such as the VRFB storage state and the load curve. The proposed solution [70] is a generalized one that declares to be valid for supplying uninterrupted power to the rural areas with weak or no grid conditions, which can ensure the probability of power supply loss down to zero at the users end. Recently, artificial intelligent (AI) optimization techniques [71] were applied to raise the life span of the battery leading to the cost reduction of the hybrid system of PV, wind generator with fuel cell and energy storage. The above-mentioned technologies and researches are summarized in Table 3.

Table 3 Renewable energy-based BESS for BEF
Full size table

4.4 Features of EV

The opinion that electric vehicles (EV) can be connected to network as mobile energy storage equipment is inspiring, particularly when considering the volatility of traditional stationary BESS, or conventional types of back-up energy supplies. EV expands the additional BESS volume. Mazzeo [73] pointed out that choosing a suitable battery has great possibility to integrate the RES with the nocturnal EV charging storage system after the analysis of energy performance, economy performance, and environment performance.

Kashif et al.[74] provides a feasible scheme for connecting EVs to power grid system. There are two cooperatives’ strategies for EV. One is a single form without cooperated with other EVs, and each EV just arranges its own charging time. The other strategy is the aggregation strategy of EVs cooperation. By designing reasonable pricing mechanism to guide owners’ behaviors, just like BES, EV charging loads can also realize valley shifting in grid and benefit to social welfare. Based on the dynamic electric price, a study [17] explored the vehicle-to-grid (V2G) market. The result shows that one EV can additionally make about $100 a month by charging and discharging.

So far, it has become a very popular research topic to study the viability and reliability of vehicle-to-grid (V2G) [75]. Basically, a V2G configuration implies that personal automobiles are possible to be not only vehicles, but mobile, self-contained resources which can administer electric flow and replace the demand for power infrastructure. Owners can drive EV when they need and use it as power sources at peak time, and then recharging it during off-peak hours such as mid-night.

The different types of vehicle-grid integration (VGI) can provide advantage to various stakeholders. For power utilities, VGI provide back-up power and help balancing load. It can also decrease peak loads and the uncertainty of hourly and daily load. Furthermore, the VGI may make greater utilization of existing generation capacity and distribution infrastructures [76]. VGI can integrate power generation from available RES into the grid since the government attempts to whittle GHG emissions down.

The appearance of plug-in battery electric vehicles (BEV) will bring a lot of distributed energy storage into the power industry. However, the flexibility potential of BEV as a distributed energy resource (DER) is limited by several factors including their mobility, the requirement to serve transport energy needs, and the locational/temporal availability of physical charging opportunities. The DER potential declines in the day time in all situations, but this reduction can be minimized by accessing to additional charging infrastructures [77]. Therefore, investment in extra non-residential charging infrastructures may be of particular importance for maximizing the DER potential of BEV storage flexibility [77].

It should be noticed that the unregulated charging of EVs may lead to an extreme expansion in electricity demand at peak hours which will have a negative impact on stability and security of grid. In order to promote widespread application of EVs, it is urgent to regulate EV charging activity efficiently [78].

4.5 EV and buildings

Over the past two decades, researchers have explored diverse concepts of vehicle-to-grid (V2G), vehicle-grid-integration (VGI), or grid-integrated vehicles (GIVs). VGI generally includes V2G, which is a vision of bi-directional flow between the vehicle and grid with a more advancing technique. Personal electric vehicle (i.e., PEV) is equal to the storage device in some way [76]. The vehicle-to-building (V2B) / building-to-vehicle (B2V) interaction can decrease the dependence of both household usage and transportation on the grid [76], as shown in Fig. 8. The main purpose is to neatly adjust the operation of both the HVAC system and the EV battery sustained by a renewable system in a building to meet the requirements of the building occupants and the grid [20].

Fig. 8
figure 8

Vehicle-to-building (V2B) concept [20]

Full size image

The advantage of EV is that, besides the role of batteries for buildings, it can assimilate residual energy produced by RES and enhance the utilization of base-load power plants by implementing smart charging plants with smart grid infrastructure. Furthermore, V2G and V2B applications can reduce peak demands. Thus, the larger scale integration of EVs in the future will have positive impacts on the power system, though it will raise transmission and distribution losses to a certain extent [79]. However, it needs the local tariff or regulations to prevent instability, in case of excess cars supplying internal storage for V2G [80].

According to the study by Chen et al. [18], shifting PV and valley electricity through V2B can not only increase utilization, but also acquire impressive revenues. Using V2B to transfer PV can gain more economic benefit than the valley electricity. The V2B energy arbitrage falls with the growth of the EV driving distance. Utilization of EVs to extra PV power storage can re-distribute energy into buildings with high demand, such as Time of Use Tariff (TOUT) or triad periods [81]. Different from investigating EV with a single building, a case study in Sweden [82] shows that the daily renewable self-consumption rate in cluster-level increases by 19% and the daily electricity bills decreases by 36% when compared with the traditional control.

Many efforts are devoted to realize net zero energy building (NZEB) by involving EV in the system. The hybrid system integrating V2B with ground source heat pump or district heating in a NZEB design may result in cost savings by 38% or 11% per year at most, respectively [83]. In another study, an advanced form of building to vehicle to building concept (V2B2) is suggested [84]. This new concept exploits for the sake of speeding up the growth of a new zero energy mode, in which the EVs are treated as energy carriers to transfer power among buildings [19]. At the meantime, this system may further promote the allocation of RES and accelerate the popularization of EVs. The energy saving percentage for an investigated battery swap option layout may vary from 38 to 73% [84]. A neoteric control strategy, called “boundary expansion scenario”, was proposed to realize NZEB with EVs [85]. It was shown that the renewable energy can cover EV storages with almost full percentage (the renewable energy ratio at 96.9%) if the matching capability among PV, wind turbine, and EV can be enhanced [85].

In the situation of power to buildings, the procurement and billing arrangements adopted by the tenants or building owner are meaningful. Commercial buildings operate according to several typical payment kinds, such as Time of Use Tariff (TOUT), fixed rate tariffs, and triads. A study [81] suggested that it has possibility for V2G electricity trading in markets. Nevertheless, the interaction among building demand, vehicle use, and market requirements is undefined. Little information about the potential, local self-consumption, or the capacity market of EVs aggregation for providing into Short Term Operating Reserve (STOR) is known.

At last, when batteries are degraded to a level in an EV and ready for retirement, they still have a considerable serving life and can be re-used in other places like energy storage unit in a micro-grid of the residential with PV system [86]. This means that batteries eliminated from an EV can still be used in building systems. A short summary is provided in Table 4.

Table 4 EV based BESS for BEF
Full size table

4.6 On the choice of BESS

There are many different types of BESS in terms of their material features, limitations and cost, and they may be applied in different fields and areas. Constraint by the content, the details are not displayed in this subsection, but it is referred to review study by Hannan et al. [90] in which the overview of different battery energy storage technologies considering life cycle, efficiency, power and energy density, advantages, limitations and applications are clearly stated, listed and compared. In addition, considering the importance of EV as a type of BESS, it is referred to study by Rezaei et al. [91] in which a comprehensive comparison of power density, energy density, efficiency, response time, cost, merits, etc. among Li-ion battery, Ni-MH battery, UC (Ultracapacitor), FESS (Flywheel), SMES(Superconducting Magnetic Coil) for EV are offered.

5 Discussions on challenges and opportunities

5.1 Dealing with battery and building aging

Every system is combating with its aging and efficiency degradation during operation. The life span is limited but the only thing we can do is to prolong its life span or keep efficiency degradation a minimum within a certain life span. The same principle also applies to the study of building energy system with BESS.

There are many studies putting forward different strategies by considering battery aging issues. The study by Wankmüller et al. [92] finds that by introducing the penalty cost into the objective function, the energy arbitrage optimization model can reduce the cycling of the battery. And the profitability of the BESS in the whole life can be improved. Cai et al.[93] presents a model predictive control (MPC) method on the basis of aging-aware for sustainable buildings operation with on-site PV and BESS, which can decline the operating cost by 9% and by 4% at most under the condition of moderately increasing battery degradation. The experimental results by Wang et al. [32] show that the battery has a promising payback time, through the optimally deep discharging rather than the common practice of shallow discharging to prevent phenomenal degradation. Compared to the battery only system, Li et al.[94] proposed a hybrid energy storage system (HESS), which consists of the superconducting energy storage system (SMES) and the battery, has directly raised the life span from 6.38 years to 9.21 years.

Researchers are calling for the advancements in battery material science to tackle the battery aging issue from the bottom. But that process could be long, and what currently we could contribute in building sector is the optimal control and management of BESS when serving with BEF. In addition, it is noteworthy that most of literature only focused on the issue of battery aging while overlooked that within a certain period of time, the building itself as well as the other energy equipment in buildings are degrading simultaneously. Especially, some studies [95, 96] start to notice that the impact of building envelope for the energy flexible buildings is substantial, which means the aging of building envelope will also play an important role.

5.2 System optimizations with multiple evaluation indexes

It is hard to set a clear and ultimate goal for the energy flexible buildings because their performances are linked to many different evaluation indexes. Most of studies only touch a few of them. There are generally four categories of indexes used in current studies.

  1. 1)

    Technical indexes: simple payback analysis, grid independency, etc. The proposed solution of PV and battery systems taking 8 years for payback makes it seems to be a questionable investment, although it saves £300 k in terms of 5-year NPV [97]. Overall, the integrated technology selection and operation (TSO) optimization model provides a valuable viewpoint into practical project evaluation, and it is conducive to decrease the non-determinacy of capital-intensive project related system. Thus, for the distributed energy technology assessments, it is proved to be an excellent modelling framework.

  2. 2)

    Economic indexes: replacement cost, electricity bill, net present value, etc. It is proved that by load serving entities (LSEs), the shared ESS can significantly reduce the peak power besides lowering the energy costs of consumers for their main objective in-home applications [98]. Nevertheless, decline of battery costs and supportive policies would confirm that batteries play an increasingly vital role in solar plus systems. Moreover, if taking all the incentives into account, the combination of energy storage and PV power generation may have more profits than PV alone system [99]. The study of Heine et al.[100] found that adding batteries which meet 1.5–1.6 times of the annual peak energy purchase demands, offers the maximum NPV for PV-battery systems in locations with appropriate utility ratios.

  3. 3)

    Tech-economic indexes: The analysis from Bagalini et al.[101] implies that PV system alone is not economically feasible because of low electricity prices, high technology costs. However, by considering the doubling of future electricity prices, the reduction of technology costs (- 66% of battery and—17% of PV and inverter) and the halving of export tariffs, PV-battery system brings more benefits and shows greater adaptability to future conditions than PV system alone.

  4. 4)

    Environmental indexes: reduction in carbon dioxide (CO2) emissions is included. Considering future incentive revoke, electricity system decarbonization and technology costs, Jones et al. [102] used discounted cash flow analysis and life cycle assessment method to evaluate the financial and CO2 influences of adding BESS to a PV assemblage. According to the Paris Climate Change Agreement, the PV and battery system decreases the CO2 emissions by 17% (19t CO2) compared with the reference over a 30-year span with grid-only system. In the study of Zhang et al. [103], the decrease of carbon dioxide emission (CDE) from PV generation is studied. It implies when a high carbon credit is available, it is possible to further shorten the payback period (PBP).

  5. 5)

    Social indexes: the design, construction, and operation of BEF and BESS in community or larger scale should also include consideration of its social impacts. Although this is a very important element, there can be hardly found a study discussing this factor related to building and energy systems, especially BEF with BESS.

However, if the multiple indexes are included together, we may find some conflicts among them. There is no ultimate optimal solution for sure, but there should be a specific optimization for a specific problem or target. It is still unsolved that how those indexes should be considered and balanced in different scenarios with each other. If it can be properly solved, it can further promote the work of building energy flexibility labelling [95].

5.3 Higher level of complexity implies higher flexibility

The building energy flexibility is facilitated by integrating with battery storage which can offer a satisfying coordination between supplier and consumer, uncertainty of renewable energy sources. But there are still much room for improvements, as recently an increasing attention is given to expand energy storage from pure electric battery to electricity-thermal hybrid energy storage. This is exhibited in two aspects: (1) adopting a battery and thermal storage coupled system; (2) recognizing the building itself as a thermal energy storage system.

For the first category, Lizana et al.[104] showcased that the end-user’s electricity bill can be saved by 20% and retailer’s associated electricity expense is decreased by 25% with the aid of a better demand response management [17] and latent heat storage in buildings.

For the second category, Vivian et al.[105] noted that the buildings thermal inertia has the potential to modify their thermal load pattern. Another case study on a single-family house and apartment block was conducted by Foteinaki et al.[96] in which they emphasized the importance of building envelope considering its thermal capacity. Quantitatively, Ramos et al.[106] pointed it out that after DSM on building thermal mass is implemented, for heating and cooling, it saves the cost by 3.2% and 8.5% at most respectively.

However, few literatures make collective consideration and deep integration of the above mentioned two aspects with demand side management (i.e. DSM) towards BEF. There is another challenge that if our studies on BEF is expanded from analysis on a single building to a building cluster, an even higher possibility maybe reached. Walker et al.[107] pointed out that the different thermal load profile among different buildings could offer a great opportunity of energy flexibility.

Besides, an even higher level of building energy flexibility could be released through integration with many more relevant systems, such as building plus EV [20] or even including regional heating system [108] and with both fixed and variable feed-in price components [109].

5.4 Distributed energy control for distributed energy systems

It is noticed that most of studies on building energy flexibility designed a centralized coordinated energy management system or algorithm [110] to better control each part of the system. This is a very common way of control and management. But we also should notice that the treatment of BEF is similar to the function of distributed energy systems. Actually, in some studies, they indeed make distributed energy system in the framework. Concerning an advanced control, multi-agent control (or distributed energy control as a counter part of centralized coordinated energy management), may offer a new horizon for the BEF study. In order to support distributed decision, a general multi-agent control methodology was suggested [111], which can be available for the “plug-and-play” of building energy system optimization. This control method can find approximately the optimal solutions and achieve remarkable energy saving effect. There are many other studies dedicated for the microgrids and building energy management [112, 113], but few applied this to the issue of BEF which leaves some rooms for future studies.

5.5 Building energy pattern dance with users’ behavior pattern

We should notice that all those components are closely linked to users’ behavior pattern. In a larger picture, the building energy pattern is closely linked to users’ behavior pattern. The way how we use energy largely determine the intensity and duration of building energy profile.

In a recent questionnaire-based analysis study in Denmark by Li et al. [114], the result is sort of surprising that over 60% of the respondents had never heard of smart grids. But on the bright side, half of the interviewees (20–29 years old) knew the concept, and 11% were found to be potential users of flexible building. In another study [115], a data-driven random occupancy model was adopted which can describe the randomness of occupancy patterns when building clusters are involved. However, we find that the researches considering users’ behavior pattern related to BEF are too limited. We think the study focus should be moved to this area, so as to realize a better application in the near future.

5.6 Critical technologies to facilitate building energy flexibility

According to the description in the previous sections, researchers also propose some implementation methods of BEF such as IoT and smart meters according to different load characteristics.

In terms of data exchange, deployment and commissioning time in small and medium-sized buildings, IoT is a good choice which can further realize the connection between individuals. In the study of Tanasiev et al.[116], to expand the scale of IOT monitoring solutions and reduce deployment costs, some concessions need to be made in accuracy. The proposed approach will lead to reduction in greenhouse gas emissions and also in overall reduction of power and heat usage. Network equipment has the potential to bring breakthrough innovation and essentially catalyze the smart city into urban life [117].

The traditional grid with smart monitoring and management equipment provides a key solution for power generation and distribution. The smart metering infrastructure proposed by Kabalcı and Siano [118] are targeted for monitoring the generation and consumption values of RESs, ESSs, DC/AC distributed generation sources and various type of loads available in microgrid and smart home environments. This kind of smart meter makes power consumption behavior more reasonable constrained and planned.

Besides, in order to better perform system control, design, and optimization, an accurate, efficient and flexible model of BESS is needed. There are some latest review studies targeting this topic which could provide technical references. Vykhodtsev et al. [119] offered very detailed modeling methods for techno-economic analyses. In addition, if the connection between BESS and renewable energy modeling is considered, Yang et al. [120] provided comprehensive reviews and suggestions. If the modeling difficulties cannot be properly solved, the latest data-driven method [121] might offer new paths to solutions.

6 Conclusions

The building energy consumption intensity and energy efficiency are two key factors frequently studied and concerned. In recent years, building energy flexibility (BEF) has been a third important element for building energy evaluation and management. Meanwhile, the integration of battery energy storage system (BESS) will playing a big role. There is a noticeable increasing research inputs in this topic while lacking of comprehensive review study. This study provided a comprehensive review and in-depth analysis on BEF with BESS, covering major aspects from some basic structures to system modeling. The following findings and research outlooks are discerned to spread the application of BESS in BEF.

  1. 1)

    Based on the data survey and statistical processing, it is found that the publication for BEF is increasing in an exponential curve. It is worth mentioning that the publication number from the top six countries possess more than half of the total literature about BEF. China plays an important role in study field, ranking second in the list.

  2. 2)

    The energy characteristics of various building types are investigated which can offer a better understanding from the angle of end users. This is an initial basis for demand response (DR) management in which the classifications and tools for DR are summarized and their applications for BESS in buildings are discussed. In the utilization, the selection of DR strategy mainly depends on the fluctuation of energy sources, the characteristics of BEF and the local power tariff policy.

  3. 3)

    The BESS can be divided into stationary (conventional battery) and mobile (electric vehicle) types. For the stationary type, the integration of various renewable energy sources including solar energy, wind energy, and PV-wind energy sources are discussed. For the mobile type, the EV offers new solutions to building energy management, but their interaction and coordination is much more complex than the stationary type. It brings a greater possibility to BEF to meet the demand of market and users.

  4. 4)

    Finally, the challenges and future research opportunities are identified from 6 aspects, including: the synchronized aging of building and battery and complex optimization; system optimization with multiple index covering from energy to social indexes; the abundant opportunities of BEF behind the challenges of higher system complexity; the great potential of distributed energy control for distributed energy system; the co-working of building energy pattern and users’ behavior pattern; as well as critical technology advances to facilitate BEF. This study may further facilitate the BEF with BESS and push forward the synergistic improvement of building energy efficiency, intensity and flexibility.