Introduction

Background information

An oversized system can be reliable, but costs so much that renders a project uneconomical. Likewise, an undersized system may be economical but tends to be vulnerable to overstretching demands that lead to early failures. Hence, there exist conflicting objectives in-between reliability and economy to be balanced. The main objectives in renewable energy (RE)-based distributed systems are reliability, energy costs, supply availability and emission control. Balancing the foregoing techno-economic objectives in distributed grid planning and operations require defined energy management (EM) schemes at both sides of the metre (the system’s front-of-the-metre and behind-the-metre). The foregoing are issues with EM schemes applied to the supply side (power management) and demand side (load management) in the emerging utilities, the distributed grids (DGs). Demand side management (DSM) is a systematic management of customer demands based on the supply availability (power dispatch) applicable to buildings and, by extension, the general real estate illustrated in Fig. 1. Power dispatch in conventional grids (CGs) is strategic to system operations. However, power dispatch in DGs may be more strategic due to the impacts of uncontrollable RE resources and diversified technologies involved, as illustrated in Fig. 2. Factors of climate change affect both the RE resources used for power generation and control in DG utilities and the customer demands such as the seasonal needs for lighting and space comforts. Thus, reliability challenges are expected in DGs such that what may matter most to the system operations is the optimal implementation of demand side management (DSM). The DSM is known to be flexible enough to be made to follow supplies in the events of intermittent power dispatch and enables customer participation.

Fig. 1
figure 1

Examples of components in a real estate

Full size image
Fig. 2
figure 2

Structural distinctions between conventional grids and distributed generation

Full size image

The implications of DSM are primarily attributed to the customer side of the electric power distribution networks (PDN). The customer side of the PDN simply comprises electricity demands from the industrial, commercial and residential settings within buildings as components of the real estate establishments. Furthermore, real estate comprises natural or man-made structures that are permanently tied to the land, which includes rivers, vegetation and mineral deposits. Man-made structures can be buildings, roads and bridges. However, the real estate focus of this study constitutes mostly of buildings, which are known to largely contribute to global electricity consumption. In contrast to intangible assets like stocks and bonds, real estate consists of physical land, buildings and other structures. The scope of the rights attached to real property (i.e. the right to enjoy, occupy, use and transfer) is governed by the legal and political procedures that have jurisdiction over the property (Friedman et al. 2017). From a scholarly point of view, the real estate field is an overarching discipline that includes subfields specialising in different aspects of the property development process (Jagun 2020). To be clear, residential real estate refers to either single-family homes or multi-family complexes, depending on the context (Cetin and Novoselac 2015). The most common sort of estate is an area where most people have some level of experience. Residential encompasses many building types, including single-family homes, multi-family complexes, condominiums, townhouses and other housing types.

In modern electricity DGs, buildings play a significant role in shaping city electricity generation, distribution, retails, and utilisation, known as prosumption (Dahiru 2021). The role of buildings in the control of greenhouse emissions is evident (Hu et al. 2020). Moreover, electricity is the most commonly used energy source in buildings, and the demands keep increasing. Hence, offsetting the implications of rising demands requires dynamic building energy efficiency strategies at the front-of-the-metre, otherwise referred to as load management or DSM. Building energy management systems (BEMS) enable estate managers and owners to increase energy efficiency in buildings for a reduction in the use of energy. Depending on the nature of the building, several management solutions are utilised for energy savings through the DSM schemes (Mariano-Hernández et al. 2021). There is a need for a BEMS to be designed to respond to the electricity grid conditions for improved efficiency and sustainable energy consumption in buildings (Al Dakheel et al. 2020). The inelasticity in building electricity consumption and stochastic human behaviour incorporated into the grid operations are current global challenges. Hence, buildings must be able to adjust electricity consumption in response to realistic market signals (Farrokhifar et al. 2021). Moreover, it is possible to achieve a more connected and efficient grid operation with the current advancement in smart homes and grid technology in buildings, which account for the larger portion of electricity consumption. The foregoing is a crucial step towards a smart grid implementation (Babar et al. 2020). Integration of RE technologies and energy storage systems (ESS) at load centres is one of the most important features of the smart grid. Thus, having RE system scheduling (power dispatch) and smart control of home energy consumption systems (DSM) based on peak and off-peak periods is suggested (Ma and Li 2020).

Literature review

The concepts of DSM are focussed on achieving a utility system’s balanced operations (Atia and Yamada 2016). The DSM may be viewed as the implementation of load management at the customer side of the DG utilities according to real-time supply availability. Objectives of DSM implementations in DGs include energy cost reductions resulting from increasing demands and prevention of early failures due to overstretching demands. Such techno-economic goals involve concerns for state policies, regulatory agencies, system operators, utilities and customers. DSM, in a broader concept, consists of demand response (DR) programmes and energy efficiency (EE) (Masters 2004). The DR is a utility-based designed programme for the short-term management of customer demands. The DR programmes provide opportunities for customers to participate in electric grid operations through shifting or reduction of electricity usage for time-based energy rates or financial incentives. Customers are attracted to respond to DR programmes through offerings such as time-of-use (TOU) pricing, critical-peak pricing (CPP), real-time pricing (RTP) and critical-peak rebates (CPR) (Masters 2004). This is essential when considering the BEMS and its strategies for RE efficiency towards real estate development and management, which include residential, commercial, agricultural, industrial and institutional buildings. The importance of emphasising energy management and efficiencies in real estate construction, particularly in building systems, is made by Calvillo et al. (2016).

It is indicated by Debnath et al. (2017), Eze et al. (2016) and Ioakimidis et al. (2018) that DSM is implemented based on clearly defined strategies sketched in Fig. 3. The traditionally used DSM strategies in the literature concerning the applications in RE-based DG systems for load management include peak shaving, valley filling, load levelling and load shifting. Peak shaving considers customers’ load shedding as initiated by utility-based DR programmes to release stress on supplies. On the contrary, valley filling, as initiated by utility-based DR, tends to raise demands against excess generation to reduce rates of energy curtailments. Load levelling is a DSM strategy needed where large fluctuations occur on PDN. Whereas load shifting considers the criticality of demand and supply availability to transfer demands among customers or appliances. Uddin et al. (2018) extensively review peak shaving as a DSM strategy concerning the integration of energy storage systems (ESS) and electric vehicles (EV) with the main grid. The traditional DSM strategies, the peak shaving and valley filling are achieved by Ioakimidis et al. (2018) to optimise power consumption profiles in a university building by scheduling the charging/discharging process of an EV parking lot using real-world data of power consumption and parking lot occupancy. Agamah and Ekonomou (2017a, b) developed an algorithm that uses demand profile information and a minimal set of ESS parameters in obtaining an ESS-based operations schedule for demand peak shaving and load levelling. A peak shaving strategy is achieved by Buja et al. (2017) to investigate the capabilities of vehicle-to-grid (V2G)-enabled EVs in executing reactive power compensations through either of the two topologies of a bidirectional battery charger. A valley filling strategy is implemented using rule-based energy management for electric charging from a photovoltaic-to-grid (PV2G) system (Bhatti and Salam 2018). Potentials of load shifting were forecasted to have reduced German-Austrian nationwide electricity expenses by 6% when a 25% DR adoption rate was achieved (Märkle-Huß et al. 2018). Other strategies in DSM include energy arbitrage, strategic conservation, strategic load growth and flexible load scheduling. Energy arbitrage is a DSM strategy achieved by energy savings during periods of lower energy costs against periods of higher energy costs. Energy arbitrage is mainly achieved using energy storage devices such as batteries, supercapacitors and EVs. A study by Metz and Tomé (2018) used mixed integer problems to optimise storage dispatch for energy price arbitrage considering time-based differential energy auctions. Likewise, Salles et al. (2016) took advantage of electricity price volatility in 7395 different electricity market locations to improve revenue using energy arbitrage of a generic ESS. Flexible load scheduling implemented using improved TOU pricing methodologies in nanogrid operations achieved a range of 1.72–53.09% consumption cost reduction under binary battery operations (Tijjani et al. 2021).

Fig. 3
figure 3

Graphical illustration of demand side management strategies for optimal load scheduling

Full size image

Strategic load growth, strategic conservation and flexible load scheduling are mentioned to be part of DSM strategies discussed in existing literature such as Jabir et al. (2018), Gaur et al. (2017), Lokeshgupta and Sivasubramani (2018), and Jacob et al. (2018). Like the traditional DSM strategies, strategic conservation is a utility-initiated programme that focuses mainly on customer interests in accepting incentives for reduced energy use. Strategic load growth is a planned increase in energy sales to improve customer productivity while increasing utilities per kWh energy sales. Whereas flexible load scheduling is a programme where customers receive incentives for load building and curtailments. In flexible load scheduling, incentives for load growth and decay are applied to customers interchangeably.

Study contribution and paper organisation

The foregoing DSM strategies are not prominently used in literature for applications despite their flexibilities and suitability to RE-based DG system problems. This paper analyses literature implementations of the DSM strategies as it applies to emerging DG frameworks and building energy management structures (BEMS). It is indicated in Fig. 4 that the reviewed literature in the areas of DSM and its relationships to BEMS have other major issues of discussion in the paper such as distributed generation/grids (DGs). Classifications of the DGs such as the traditional CGs (in macrogrid and minigrid architecture), the emerging microgrids and nanogrids are extensively discussed. Other related literature consulted and discussed in the paper include optimization techniques, renewable energy (RE) and the impending issues of energy cost. Hence, the paper contributes to the identification of the following points concerning existing literature applications of DSM strategies in DG system developments and building energy management, as highlighted in Fig. 4.

  • The DSM strategies implemented based on utility-initiated demand response (DR) programmes are real estate oriented, which covers residential, commercial, institutional and industrial buildings.

  • Dump energies due to power generation curtailments are not as economical as it appears. The process may incur additional costs and power losses.

  • The real estate’s contribution to energy conservation and emission control could be targeted through the development of nearly zero energy buildings (NZEB).

Fig. 4
figure 4

The review’s related areas discussed in the existing literature (VOSviewer)

Full size image

The rest of the paper is arranged by providing an overview of emerging DG systems based on the perspectives of real estate in the “Distributed grid technologies” section. System planning and operations as the main aspects of the development of DGs are discussed in the “Classifications in distributed grids” section. A detailed review and overview of DSM strategies discussed in the literature is presented in the “Distributed grid planning and operations” section. State-of-the-art applications of DSM strategies used in literature for the development of modern and smart DG systems are discussed in the “Demand side management and strategies” section. An overview of literature perspectives concerning energy management in buildings is discussed in the “Other concepts related to demand side management” section. Implications of climate change on the performance of RE resources and technologies, on the one hand, and seasonal change in customer demands, on the other hand, are discussed in the “Implications of climate change to renewables and energy use in buildings” section. A critical analysis of existing literature regarding research findings with recommendation for further work is provided in the “Limitations and future considerations” section. The paper is concluded in the “Conclusions” section.

Distributed grid technologies

The need for a study to focus on the development of power grids against the challenges of the twenty-first century is highlighted for the benefits of reliability, efficiency, cost-effectiveness and environmental considerations (Baek et al. 2017). The conventional grid (CG) is a vertically structured centralised system that comprises major units which include generation, transmission, distribution and retail (Ma et al. 2017). Challenges affecting the performance of CGs include the high costs of fuels and the effects of volatile oil markets, transmission losses, carbon emissions and high acquisition costs (Zenginis et al. 2017). Maintaining such systems could be costly in terms of economy and environment (Deckmyn et al. 2017). Other challenges could be reliability due to human errors, natural disasters and transmission losses (Islam et al. 2017; Burmester et al. 2017). Part of the solution to the foregoing problems is to improve supply reliability through increased generation. It is viewed that the power grid’s transmission losses can significantly be reduced by the adoption of modern DG structures, and the need for expansion of the existing structure can be eliminated (Debnath et al. 2017). The DGs are as well good for the support of RE integration to reduce carbon emissions. The DGs also ensure minimisation in the rate of maintenance in addition to fuel and energy consumption cost reduction (Ganesan et al. 2017).

Apart from the significant impacts of DGs in power systems’ support for RE generation and integrations discussed by Tudu et al. (2019) and Kuang et al. (2016), the REs are good in the harvest of free, abundant and lifetime resources. REs also enable low-capacity generations (microgeneration) for customers to optionally and simultaneously consume and produce electricity in an energy trade-off framework termed “prosumption” (Genikomsakis et al. 2017). PV cells, wind turbines (WT) and fuel cells (FC) are a few examples of RE components used as generators in emerging DG systems. The distinctive features of CGs as compared to DGs are represented in the hierarchical diagram of Fig. 5. In the hierarchy, DGs are shown to comprise microgrids and nanogrids as scale-down systems, usually designed for convenience, costs and logistics. Smart grids shown in the diagram indicate structures of DGs as units or subunits equipped with communication facilities. Such systems feature an example of preferred characteristics of a modern power grid highlighted in Table 1.

Fig. 5
figure 5

Hierarchies in modern electrical power grids

Full size image
Table 1 Characteristics of CGs, DGs and preferred modern power grids (Masters 2004)
Full size table

Classifications in distributed grids

Technically speaking, DGs are described as an assembly of different types and sizes of RE technologies such as electrical energy sources, energy storage and consumer appliances linked through power cables, power converters and power control devices as interfaces. There may be no definite topologies in DG architecture, as the composition of components in the system differs depending on planning and operational objectives. However, system capacity and energy demand requirements could be a basis for the classification of DGs concerning components and system sizes. Classifications in DG structures and topologies are usually determined by factors such as geographical location, number and size of buildings, critical demands and bus potentials in the PDNs. Hence, in DG systems capacity-based topologies, the classifications may consist of the macrogrids, the minigrids, the microgrids and the nanogrids.

The macrogrid

Macrogrid topologies are usually centralised systems designed in either ring or radial systems to serve customers within an extensively large area and large population. Generation in this topology is mainly kept at long distances away from load centres. Hence, power transfers from generation to customers are supported through transmission networks, as illustrated in Fig. 6. The architecture in macrogrid topology may be considered as the typical traditional CGs. Voltages handled in this topology are the traditional generation voltages (5 to 34.5 kV), transmission voltages (66 to 765 kV) and distribution voltages (120 to 240 V single-phase and 220/420 V to 33 kV three-phase). The topology is also referred to as the main grid, utility grid or legacy grid. This is the topology dominating larger global economies with heavy residential, commercial and industrial demands.

Fig. 6
figure 6

The basic structure of a macrogrid system

Full size image

The minigrid

Factors considered in the design and implementation of minigrid topologies include distance or isolation from macrogrid access, such that service extensions are largely affected by implications of the cost involved. The topology is smaller than the macrogrid in terms of generation capacities and servicing consumer demands. Transmission networks may not be part of this topology as the distance between generation and load centres are significantly close, such that local PDNs are adequate for power transfers. Figure 7 indicates that customers in this topology are mainly residential and commercial. However, this topology is prone to uneconomical and environment-unfriendly generating systems such as multiple diesel plants, biomass or small hydro facilities (Javaid et al. 2018). Generation capacity usually ranges between 1 kVA and 10 MVA in minigrid topologies (IRENA 2018), while it supports low voltages (120 to 220 V) mainly for distribution purposes (Program and ESMAP 2000). Minigrid topologies are mainly designed for developing economies and developed countries with remote locations such as islands. The topologies are known to have high operating inefficiencies and protection issues. Hence, the system is not usually suitable for urban deployments.

Fig. 7
figure 7

The basic structure of a minigrid system

Full size image

The microgrid

Based on the United States (US) Department of Energy (DOE) definitions, the microgrid topology is an electrical entity of generators and loads operating either in isolated mode, in connection with other grids (macrogrids or minigrids) or a network of other microgrids. The system employs either a combination of fossil fuel and RE-based generators. The microgrid topology may not have a definite size; however, the World Bank describes its operating voltages to be below 11 kV (Javaid et al. 2018). Technologies supported in microgrid topologies are highly diverse based on their types and sizes. A microgrid is a typical example of a DG system network with resources sparsely connected, as illustrated in Fig. 8. Issues affecting this topology are complex structures due to the integration of various energy sources and inadequate standardizations. A relatively small market currently affects microgrid prospects to have industry-wide technology standardizations. It is expected that proper standardizations in microgrid systems are necessary to achieve large-scale price reductions and high degrees of interoperability.

Fig. 8
figure 8

Multi-energy structure of renewable energy-based grid interactive microgrid

Full size image

The nanogrid

Nanogrid is an electrical grid topology that takes a general approach and design principles of microgrids but with much smaller capacity, lower voltage level and lower needs for optimization complexities. A key feature of nanogrid is their ability to interconnect with macrogrids, microgrids and adjacent nanogrids. In similarities with the microgrids, the underlying philosophies about nanogrid concepts are economy (reduction in energy and operational costs, elimination of the cost of macrogrid extension), reliability (increased supply availability), environment (reduced rate of power generation-based emissions) and speed (the reduced time it may take to extend macrogrid services). As obtainable in microgrids, the nanogrid structure illustrated in Fig. 9 operates utilising AC/DC sources and storage. However, it is stated by Nordman (2009) that nanogrids have no concern for a power source. The foregoing view could be considered an outdated concept as any electrical entity without a power source may hardly be autonomous and may be regarded simply as controlled loads within a network of other grid entities. Hence, an entity without a power source may not be qualified for the “grid” affixations. Justifications for the concepts of powered nanogrids are discussed in recent literature where nanogrids are configured to operate in either autonomous or interconnected operations with contributions from nanogrid’s local power sources (Akinyele 2017; Shahidehpour et al. 2017; Moussa et al. 2019; and Cordova-Fajardo and Tututi 2019).

Fig. 9
figure 9

Structures of renewable energy-based AC/DC hybrid nanogrids

Full size image

Smart grids and emerging power networks

The existing electric utility services face challenges such as the need for RE penetration, changes in market dynamics, rapid technological advancements and shifts in customer affordability and tastes while operating the grids. Thus, the foregoing challenges require resilience, accurate forecasting and security against threats both internal and external. The foregoing challenges hence attract unbundling and restructuring in the existing grids and optimising its assets. Comparisons made to distinguish between modern grids and existing grids outline the following aspects needed for the development or transformation of existing grid structures into preferred modern grid architecture referred to as smart grids (Momoh 2012).

  • Power system enhancements:expansion and use of REs to offsets the impacts of capacity requirements and carbon emissions.

  • Communications and standards: advanced automation and communication systems for the existing power systems are expected to generate a vast amount of operational data to be used in rapid decision-making.

  • Computational intelligence: the use of advanced analytical tools for network system power optimizations.

  • Environment and economy: customer participation and general enhancement of generation, transmission and distribution networks.

Smart grids in Linden et al. (2014) are defined by Kylili and Fokaides (2020) as “an electricity network allowing devices to communicate between suppliers to customers, allowing them to manage demand, protect the distribution network, save energy and reduce cost.” Computational intelligence and communication facilities are the key aspects featured in any smart grid system that are completely alien to legacy networks (traditional grids). This implies that DG topologies such as minigrids, microgrids and nanogrids are classified as smart grids only when the key aspects such as communication facilities and computational intelligence demonstrated in Fig. 10 are included in their design.

Fig. 10
figure 10

Simplified structure of a smart grid system

Full size image

Comparative information regarding various types of electricity grid architecture is contained in Table 2. The classifications in the table are made based on the topologies of the grid, the technologies employed, the voltage size handled and the opportunities for deployment. It is worthy of note that the emerging DGs are critical to developments within the real estate sector, as examples are set in the study for nearly zero energy buildings (Martirano et al. 2017).

Table 2 Electricity grid features and characteristics
Full size table

Distributed grid planning and operations

Techno-economic factors such as system reliability enhancement, emission control, renewable penetration, the life cycle cost, the net present cost and the cost of operation dominate objectives in renewable system design. Thus, achieving the foregoing objectives motivates the implementation of schemes for system planning and operations. To highlight the relevance of feasibility and best performance in the implementation of planning and operation schemes, advantages of optimization algorithms are usually taken. However, there are always complexities when handling schemes that involve uncertainties such as the RE resources. Hence, RE resources are characterised by the intermittent performance of generating components, such as PVs, WTs and storage facilities mentioned in the previous sections. This implies that optimal planning schemes are essential to the system sizing and placement objectives.

The use of RE resources in DG systems considering their resource dispersion, characteristics and intermittence not only affects the systems’ protection and control architecture but also makes it more complex to predict due to size, operational characteristics, cost implications and reliability. Apart from intermittent renewable resources, irregular customer demands contribute to uncertainties in DG performance. Moreover, the intermittent RE resources such as solar insolation and wind speeds affect DG systems’ reliability obliged alternative use of ESS from either a battery as suggested by Strnad and Prenc (2017) and Metz and Tomé (2018), supercapacitor suggested by Zhang et al. (2017) and Sellali et al. (2017), hydrogen storage suggested by Mendes et al. (2016), flywheel suggested by Abazari et al. (2019) and Li et al. (2019) and hydrostatic technologies suggested by Wang et al. (2018a, b). A combination of any two or more of the foregoing energy storage technologies in hybrid form also proved to be effective. Thus, planning and operational coordination in such a multi-source and multi-technology DG system require effective optimization techniques to achieve operational stability and reliability under minimised acquisition and running costs (Li et al. 2017). In Table 3, details of planning and operation schemes are provided to highlight the existing relationships within the hierarchy shown in Fig. 11 about the emerging DGs. The emphases given to the DSM (front-of-the-metre activity) strategies and schemes given in Table 3 furthermore clarify the flexibilities of the DSM and its preference for customer support and participation as advantages over the energy management (behind-the-metre activity) schemes.

Table 3 Highlighting the relationships among the distributed grids’ planning and operational schemes
Full size table
Fig. 11
figure 11

Schemes in DG system’s project planning and operations

Full size image

It is emphasised by Loucks and van Beek (2017) that for desired values in system designs and operating policy variables to be attained (such as a complex DG system), identification and evaluation of desired goals and objectives are desired. The goals and underlying objectives are hence achieved using optimization and simulation models. The desired values achieved by suitable optimization methods will lead to the highest levels of system performance and eliminate inferior options.

Demand side management and strategies

The main objective of power dispatch or unit commitment (UC) in a DG is the economic dispatch of generating components based on supply availability and demand response programmes. Whereas DSM’s main objectives include cost-effective load scheduling based on demand response programmes at the customer end to achieve the following (Attia 2010):

  • maintaining a load factor as close as possible to 1.0 and

  • maintaining a peak within the proper supply/demand margin.

By achieving the foregoing objectives, utilities could get adequate energy from participating generating units, thereby maximising profits and minimising the per kWh cost of energy. To that effect, traditional DSM strategies such as peak shaving, valley filling, load shifting and energy arbitrage were implemented by Debnath et al. (2017) and Augusto et al. (2017). The DSM was used to minimise energy deficits in a small-scale grid interactive DC microgrid for residential, telecommunications and data purposes (Pannala et al. 2017). Other solutions to energy deficits and stability to demand and supply fluctuations in microgrids were obtained by Molderink et al. (2010a, b). A poor response to peak loads was reported to be one of the limitations of rural RE-powered microgrids. Hence, the use of BESS and diesel plants were suggested to be a good option for solving peak demand issues. However, improper implementation of DSM was discovered to be a serious setback (Augusto et al. 2017). Operational balance was achieved by combining two control methods of prediction and scheduling (Hoogsteen et al. 2016). Primary frequency stabilisation is sought through DR programmes that deploy efficient DSM strategies (Azim et al. 2016). Affine arithmetic is a proposed model for load forecast that minimises uncertainties in demands (Avila et al. 2015). A shift from the known centralised control type of DSM was proposed by (Balakrishnan et al. 2017) using an agent-based method for efficiently balancing supply and demand. The DSM is known to be implemented based on suitable strategic schemes. An example of a peak clipping strategy applied in literature is a successful trade-off achieved by Nunna and Doolla (2013). In Amrr et al. (2018), reliability and economy-centred power management and load scheduling were activated using two control schemes of mode selector and source selector. A standalone nanogrid based on PV only was designed for a residential house using a Middle East climate. Load calculations for sizing of batteries and PVs, power consumption, control and monitoring were done by Akmal et al. (2016). Saini (2007) categorises DSM into the following three (3) activities upon which implementation strategies may be derived and discussed.

  • Energy demand reduction programmes. This is an activity where demands are reduced through better and more efficient processes such as smart energy buildings or the use of energy-efficient equipment.

  • Load management programmes. This is an activity of changing load patterns through demand shifts and demand curtailment during peak periods and peak rates.

  • Load growth and conservation programmes. This is an activity for change of load pattern through substitution or deferment of loads.

In summary, DSM is defined as a programme designed to modify customer use of electricity for energy and cost-related savings. The DSM programmes are usually prepared to control energy consumption on the customer side of the metre (behind-the-metre). Many strategies used in DSM applications are designed to reduce dump energies (energy wastages), curb energy consumptions (in shiftable loads) during periods of low electricity supplies, reduce energy costs and minimise the system’s costs of operations. There are many strategies designed or adapted in literature for effective DSM implementations. Examples of such include a study prepared and presented by Debnath et al. (2017) to investigate a wide range of DSM strategies whose demand curves are illustrated in Fig. 12. The investigated strategies were meant to be used in enabling customer participation and management of DGs as one of the preferred characteristics of emerging modern power grids.

Fig. 12
figure 12

Demand side management strategy curves

Full size image

Peak shaving

Peak shaving also known as peak clipping is a traditional DSM strategy that considers cutting off a portion of loads at a time in peak hours of electricity demand without adverse effects on overall demand curves. The benefits of peak shaving are identified by Molderink et al. (2010a, b) as a solution to varying daily electricity needs, clean energy production and additional unpredictable loads such as EVs. There are examples of peak clipping implementation in the literature. Peak clipping is achieved through the modification of a reference load by applying a cap for maximum power that can be delivered on request (Augusto et al. 2017). Simulation results by Martirano et al. (2017) show that a reduction of peak power by 12% and average power of 23% was achieved in a smart microgrid equipped with heating and domestic hot water. A selective peak clipping is proposed by Nunna and Doolla (2013) to achieve energy balance in microgrids. Battery power is optimally synthesised over a specified period to achieve peak shaving and reduction in grid energy buffering (Serpi et al. 2017). Prototype control methods were used to validate the simulation results of a microgrid connected to the grid and in island modes for peak demand minimisation and diminishing of stress on a CG network (Pannala et al. 2017). The advantages of peak shaving to power utilities, according to Uddin et al. (2018), include a reduction in per kWh electricity generation cost. Likewise, customer benefits from corresponding per kWh energy cost reduction as utilities normally transfer the economic burden to customers. The major disadvantage of peak shaving is a breach in customer comfort.

Valley filling

There are periods when low demands are experienced regarding base loads. Higher levels of power generation during such periods may comparatively be at a loss if demands are not raised due to unused (dump) energies. Considering the RE type of power sources in this scenario and the costs of energy involved, generation curtailment may likely not be an optimal decision. The use of storage and flexible (transferable) loads are examples of methods of achieving valley filling. An example of a valley filling strategy was achieved by Augusto et al. (2017) by restricting flexible loads to off-peak demand scenarios. Valley fillings were also realised in a study that proposed charging an EV from a PV/grid system using a rule-based EM system (REMIS) (Bhatti and Salam 2018). To regulate power consumption in a building, vehicle-to-building (V2B) concepts were formulated to achieve peak shaving and valley filling through mathematical modelling in MATLAB (Ioakimidis et al. 2018). Stochastic algorithms with Monte Carlo simulations were used to stabilise the grid through valley filling strategies (Nazarloo et al. 2018). A binary programme was used to achieve desired aggregate load profile through valley filling (Sun et al. 2016). A centralised charging is proposed by Liu et al. (2017) to be applied as a method of flattening the demand profile through valley filling for overnight charging. Advantages of valley filling strategies to utilities include a reduction in dump energies and the burden of generation curtailments are avoided. Customers may as well benefit from valley filling through flexible loads that are normally shifted to periods of lower per kWh energy costs.

Load levelling

In power distribution networks where large fluctuations of loads are experienced, load levelling is the most appropriate DSM strategy required to be applied. In load levelling strategy, efforts are ensured to reduce differences between the highest and lowest values of demand profiles (Agamah and Ekonomou 2017b). In literature, diverse load levelling applications were implemented as an important EM strategy for the reduction of losses and stability of DGs. Maximum power point tracking (MPPT) controller was used in [47] to improve energy storage and load levelling. In a study presented by Agamah and Ekonomou (2017a), load levelling objectives were combined with objectives of optimal ESS and peak demand schedule. Multiple benefits of energy storage were reviewed by Nikolaidis and Poullikkas (2018), which include load levelling as a DSM strategy. The load levelling was also realised through integrations of EVs into smart grid systems, where frequency regulations and other ancillary services were found to be beneficial to utilities (David and Al-Anbagi 2017). In Buja et al. (2017), reactive power-based load levelling was also implemented. Reducing energy costs and protecting consumer privacy was achieved by Chin et al. (2017). Resources such as ESSs and EVs are mainly used in load levelling implementations, thereby eliminating the need for network expansions.

Load shifting

In load shifting strategies, load shedding or load building synonymous with peak shaving and valley filling, respectively, at a particular point in time are not considered. Rather, the strategy considers the transfer of loads from one appliance’s demands to another based on the criticality or flexibility of loads and periods of supply availability. Examples of load shifting discussed in the literature include investigations into the potentials of peak shaving, valley filling and load shifting (Zhang et al. 2015). Load shifting strategies were applied in the sizing and selection of ESS for a PV/wind-powered DG distribution system (Sepulveda et al. 2018). The benefits of load shifting were analysed while optimising generator and load sizing schemes in a standalone microgrid (Akram et al. 2018). A day-ahead load shifting is considered while investigating the benefits of DSM to both the utility and customer sides of the DG. There were classifications of PV users based on pre-grids and post-grids terms, where the latter recorded energy savings through load shifting activities and sufficiency attitudes (Wittenberg and Matthies 2018). Load shifting was proven to increase levels of self-reliance in terms of energy consumption for a heat pump coupled with PVs (Romaní et al. 2018). An achievement of up to 6% reduction in nationwide electricity costs was recorded, as claimed by Märkle-Huß et al. (2018). The load shifting strategy was used in the reduction of peak demand and reshaping of a load profile in the proposed multi-objective optimization framework (Bastani et al. 2018). It is worthy of note that load shifting as a strategy is feasible only when load classifications are made based on demand’s criticality and flexibility rankings.

Energy arbitrage

Energy arbitrage is described as energy vending at the time of higher energy prices after a stored purchase at lower prices. Energy arbitrage may also be considered as efficient storage of energy during excess production for use at the time of low or loss of power supplies. This strategy is mostly suitable to REs based or RE-dominated DG technologies. Energy arbitrage is usually achieved using efficient storage systems that can be BESS, pumped hydro, supercapacitor, compressed air, hydrogen storage and flywheel (Nikolaidis and Poullikkas 2018). Examples of energy arbitrage implementation in the literature include a study presented by Salles et al. (2016) which considers reducing costs and improving the energy efficiency of generic model ESSs. Mathematical modelling is used to determine the feasibility of investing in energy storage of vanadium redox batteries, which recommended optimization to 75% for energy arbitrage to be profitable (Coronel et al. 2018). Energy arbitrage was listed among the numerous benefits of using ESS in electric power grids (Nikolaidis and Poullikkas 2018). A study presented by Cui et al. (2017) considers what was referred to as an extended-term energy storage (ES) arbitrage problem through a bi-level ES arbitrage solution. A portfolio theory-based approach was also proposed to achieve optimization of energy storage capacity share in different energy markets. Hence, frequency response and congestion costs were proposed for distributed network operator (DNO) markets (Yan et al. 2018). A sensitivity analysis was carried out to ascertain the price volatility required to generate profit from energy arbitrage operations (Metz and Tomé 2018). A community LV distribution system equipped with ANFIS was used for voltage management, energy arbitrage and peak load reduction, respectively (Wolfs et al. 2018). The increased complexities due to increasing penetration of REs in power systems were reduced by increasing storage requirements at the LV network that calls for stochastic integer linear programming for the solutions (Touretzky and Baldea 2014). Energy arbitrage using ESS may be used as dump loads against implications of excess generations. However, dump energies may persist with ESS fully charged. The cost of managing stored energy in ESS during charge/discharge actions against losses is yet another challenge.

Strategic conservation

Strategic conservation is a consumer-centred DSM strategy that usually originates from utility-based DR programmes specific to changing power usage patterns. Incentivised sales and usage reduction change load shapes in the programme. It is pointed out by Khan (2019) that strategic conservation is achieved by making efficient use of energy or by reducing the amount of energy service. This clearly defines strategic conservation of energy as an effort to reduce consumption by using fewer energy services. Although Kumar and Harish (2014) insisted that for strategic conservation and other DSM strategy implementations to be successful, a demand forecast that defines how electricity is consumed shall be prepared. It is found in the literature that a control technique is applied by Romaní et al. (2018) to reduce energy import for load shifting and load conservation strategies against peak periods. It is worthy of note that with differing customer tastes and behaviour, demand forecasts tend to be difficult.

Strategic load growth

This is a planned increase in energy sales ahead of valley filling strategies due to the utilisation of smart power technologies such as EVs, automation and industrial process heating. This strategy aims to increase the market share of loads that sometimes involve the addition of new customers. Strategic load growth programmes are aimed at improving customer productivity while increasing utilities per kWh energy sales. Strategic load growth is similar to valley filling, but the level of sales in strategic load growth is greater. A lot has been mentioned the strategic load growth in literature such as Debnath et al. (2017), Eze et al. (2016), Attia (2010), Kumar and Harish (2014), and Al-enezi (2010), and without any known research implementation undertaking. Advantages of the strategy include minimisation of dump energies and energy cost savings. The method could however be feasible only to systems with efficient dump loads such as energy storage systems and in cooperation with another DSM strategy such as valley filling and load shifting.

Flexible load shapes

This is a programme where customers receive incentives for load curtailments as a result of deterioration of reliability or quality of service. The DSM programme by the utility deviates from permanently sticking to a specific load shape, such that incentives attached to load growth and decay on the customer sides are applied interchangeably. Like strategic load growth, the flexible load shape is another DSM strategy whose concepts are largely mentioned, but without known research implementations in the literature. The traditional power system structures (CGs) may not find flexible load shape strategy as convenient for applications due to highly predictable power generation patterns that tally with the customer demands most of the time. However, emerging DGs operating under highly intermittent RE generations require smart DSM strategies such as flexible load shapes, such that customer demand curves may be shaped based on supply availability, energy cost and customer demand fitness (energy cost affordability). Hence, the comforts of customers need to be pegged on the fitness functions of every individual customer (Tijjani et al. 2021). The flexible load shape strategy is envisaged to improve the autonomy of a DG system interacting with the main grid. However, the flexible load shapes may not be suitable in standalone systems where unified tariff regimes are applied.

Other concepts related to demand side management

It is established that the techno-economic objectives in energy systems projects cover both planning and operation schemes. The operation schemes mainly consider EM strategy applications for optimal power dispatch, storage management and the application of DSM strategies. Power dispatch is strategic to energy system operations. However, power dispatch may be more strategic in the operations of an RE-based DG architecture due to the diverse technologies and intermittent resources involved. Hence, consequential reliability challenges are expected in DG operations such that what matters most to the system’s optimal operations are implementations of DSM strategies. The DSM is known to be flexible enough to be made to follow supplies in the event of intermittent power dispatch. In the implementation of power dispatch, storage management and DSM schemes, optimization tools and methods are found to be mostly useful. Optimization as a concept is designed to minimise or maximise the output parameters of a system by optimal selection of input parameters (Insam 2017). Achievements in realising the main techno-economic objectives which include capacity reduction, cost minimisations, profit maximizations, energy consumption reductions, emission mitigations and efficiency enhancements emphasise the importance of optimization methods to the initialization and management of energy system projects.

Optimization concepts and algorithms

Optimization is a household name in almost all fields of human endeavour, engineering, business development, industrial activities, internet routing and holiday planning. The main focus of optimizations in energy systems covers minimisation of costs, reduction in energy consumptions, time resource management, maximisation of profits, increased outputs, improved performance and better efficiencies (Yang and He 2016). Energy system applications are highly analytical that demand mathematical programming in finding solutions to real-world problems. In energy system optimizations, computer simulation tools are used in either user-designed software or commercially available off-the-shelf application software. Whatever software is considered for implementations in energy system optimizations, an algorithm or step-wise guidelines define the method upon which problem solutions are achieved.

$$\mathrm{Minimise}\; {f}_{i}\left(x\right), \;\;\;\;\;(i=\mathrm{1,2},3\dots \dots \dots \dots \dots \dots M)$$
(1)
$$\begin{array}{cc}\mathrm{Subject\; to}\; {h}_{j}\left(x\right),\;\;\; (j=1, 2, 3\dots \dots \dots \dots \dots \dots N)\end{array}$$
(2)
$${g}_{k}\left(x\right), \;\;(k=1, 2, 3\dots \dots \dots \dots \dots \dots P)$$
(3)
$$\mathrm{where}\; x=({x}_{1}, {x}_{2}, {x}_{3}\dots \dots \dots \dots \dots {x}_{d})$$
(4)

Optimization is defined by Insam (2017) as a method of economic selection of inputs for a system or process to achieve the best and yet feasible outputs based on underlying techno-economic constraints. Optimization must have a carefully defined algorithm for its implementation. In mathematical terms, most optimization problems are written in generic forms given by Eqs. (1)–(4), where \({f}_{i}\left(x\right)\), \({h}_{j}(x)\) and \({g}_{k}(x)\) are functions of the design vector and \({x}_{i}\) of \(x\) are design variables that can be real continuous, discrete or a mixture of the two. The choice of optimization algorithms depends on what problem is on ground to solve. Hence, Yang and He (2016) stresses that the following options and relevant questions need to be understood for a proper choice of optimization algorithms to be meaningful.

  • For a given type of problem, what is the best algorithm to use?

  • For a given algorithm, what kind of problem can be solved?

The two statements and corresponding questions may not be easy to comprehend as they sound. However, it is important to understand that many problems can be solved using a particular algorithm, as there are many efficient algorithms to use in solving particular problems. It is worthy of note that the choice of algorithm largely depends on the expertise of a designer (decision maker) based on the types of the problem on the ground and available resources. Other factors that may be considered in the choice of optimization algorithm are computational costs, software availability and time constraints for a solution to be achieved. In Urbanucci (2018), optimization concerning technology-based applications is implemented in a three-layered hierarchy. The synthesis, design and operations are shown in Fig. 13.

Fig. 13
figure 13

Optimization layers considered in the DG project

Full size image

Demand response programmes

The main objective of DSM in RE-based DGs is to, as much as possible, maintain a real-time balance between energy production and customer demands. Thus, DSM implementation aims to reduce energy costs resulting from increasing demands that also overstretch existing utilities. Such energy cost reduction goals include factors that involve concern for governments, system operators, utilities and customers. Other factors include environmental objectives such as emission reduction. The DSM, in a broader concept, consists of demand response (DR) programmes and energy efficiency (EE). DR is described as a designed programme for the short-term management of energy demands on the customer side of the electricity network. The DR provides an opportunity for customers to participate in electric grid operations through shifting or reduction of electricity usage for time-based rates or financial incentives. The DR programmes are designed to lower retail rates. Customers are attracted to respond to DR programmes through offerings such as time-of-use (TOU) pricing, critical-peak pricing (CPP), real-time pricing (RTP) and critical-peak rebates (CPR). The DR programmes also provide the opportunity for direct load control of heavy appliances such as air conditioners and water heaters. In preferred modern power grids, automatic switching is employed in diverting or reducing power in strategic places to avoid overloads and power failure. Advanced metering infrastructure is used to expand the range of time-based rate programmes. These programmes also have the potential of reducing peak demands that save utilities from power production and management costs, where expansion and extension of existing infrastructure are deferred.

Standard residential tariff system

In standard residential tariffs, billings are prepared based on monthly kWh consumed in addition to metre and equipment charges. The kWh electricity is the actual energy consumed by customers, whereas metre charges cover expenses incurred while installing and maintaining electricity metres. There are two structures upon which standard residential tariffs are formed. The inverted block rate structure refers to billings based on tier blocks. The billing system considers the initial tier with lower rates for kWh consumed initially within a certain tier block. The subsequent tier blocks are charged with rates higher than the preceding tier blocks. This structure of standard residential tariff is designed to discourage excessive consumption for energy conservation purposes. The other tariff structure is based on declining block rates. The structure is opposed to energy conservation and makes electricity cheaper as customer demand increases. The two structures are usually applied interchangeably to manage the usage of electricity during the summer and winter seasons.

Residential time-of-use tariffs

Residential time-of-use (TOU) tariff structures are designed to persuade customers to shift loads away from peak demand periods. Demands for electricity usually increase based on seasonal factors. During summer seasons, demands are raised in the afternoon for loads such as highly rated air conditioners for space comforts. Peak demand periods during winter are experienced for space heating. Hence, such periods of higher energy demands are managed by the application of TOU rates such that more charges for electricity bills are recorded. In other words, TOU rates/charges are much higher for electricity usage during on-peak periods. An example set by Masters (2004) indicates that TOU rates are more costly than the standard rates. Based on the foregoing examples, conventional TOU resembles standard tariff’s inverted block rate and declining block rate structures, with a more adverse effect on customer interests. In other words, conventional TOU is more economically favourable to the utilities. However, TOU provides opportunities for customers to generate, use and export electricity to utility systems through net metering. Customers with roof-top PVs have opportunities to escape the need for expensive utility-supplied on-peak electricity. In this case, monthly net electricity consumed or generated is billed or credited to the customer at applicable TOU rates.

Real-time electricity pricing

Real-time pricing (RTP) is described as the ideal structure for electricity pricing where the true cost of electricity is reflected in rates that constantly change throughout the day on a daily basis. The RTP structure of electricity pricing is offered by utilities based on day-ahead, hour-by-hour real-time pricing. This proposed tariff structure is viewed as an improvement over the TOU structure that attempts to capture the true cost of utility service. The TOU is viewed to be relatively crude since they only differentiate between a structure of a large block of periods such as on-peaks, off-peaks and partial-peaks. The RTP may require advanced smart metering equipped with effective communication facilities. RTP may also swing larger customers’ use of electricity periodically, a case that may lead to highly overstretching demands at one time and high dump energies at the other.

Implications of climate change to renewables and energy use in buildings

The RE resources are certainly dependent on climates, which vary significantly with time and region. The climates are known to be susceptible to changes, where methodologies in existing literature were used to assess or estimate the implications on RE performance and customer demands such as the smart buildings. This motivated a review to assess projects’ quantitative estimates of climate change that affect RE technologies such as solar, wind, hydro, biomass and fuel cell. The effort was to address the economic estimates and value chain based on existing gaps within the RE technologies and certain geographical regions (Solaun and Cerdá 2019). Errors were observed in RE capacity (planning) and production (operations) among the United States and European countries, where attractive policies and further research were suggested to achieve reliable RE technology and accurate weather predictions (Al Irsyad et al. 2019). Gernaat et al. (2021) considered an integrated assessment model to estimate the effects of climate change on eight renewable technologies across warming scenarios.

In addition to the implications of climate change on the planning and operations of the RE supply system, the demand side of the utility is also affected. Buildings, as examples of the demand side entities, require designs for efficient energy utilisation. Hence, Cabeza and Chàfer (2020) systematically explored technological options and strategies towards zero energy buildings. Zhai and Helman (2019) objected to the existence of several models for the assessment and mitigation strategies on climate change without a narrow prediction of its influence and identified four models accurate enough to predict the potential energy implications of climate change in a campus building stock. A statistical and dynamical down scaling method for the investigation of the energy demands for space comfort in buildings was utilised (Berardi and Jafarpur 2020). Flores-Larsen et al. (2019) evaluated the impacts of climate change on the energy performance in residential buildings by simulating a compact mid-income house using EnergyPlus. The software was used to analyse whether bioclimatic strategies were appropriate for the design of future buildings.

Building energy management systems (BEMS) are currently utilised to manage electricity utilisation in buildings. BEMS is defined as a set of tactics and procedures used to increase the performance, efficiency and energy utilisation in a system (Bonilla et al. 2018). BEMS approaches are divided into two categories, the active and the passive. The passive approach is based on offering future strategies and increase in the user’s energy awareness to indirectly influence and minimise the consumption of energy in buildings. The passive approaches are sometimes achieved using incentives. Active techniques are based on a combination of the actuators and sensors infrastructure that exists in a structure. Using smart building actuators and devices to regulate energy waste scenarios, they rely on lowering energy waste levels in their environments (Degha et al. 2019). It is worthy of note that communication facilities are always essential in the implementation of active approaches.

Limitations and future considerations

Reliability, stability and economy are the broad objectives usually attained in optimal DG system planning and operational design goals, as summarised in Tables 4, 5, and 6. The major classifications in DG operations were the power dispatch and the DSM. While appreciating the advantages of the DSM over power dispatch in RE-based systems, Kylili and Fokaides (2020) emphasise better DSM implementations through flexible load management. In recent studies, traditional strategies such as peak shaving, valley filling, load shifting and energy arbitrage were implemented, as exemplified by Augusto et al. (2017), Hossain et al. (2018), Martirano et al. (2017), Yaghmaee et al. (2017) and Fernandez et al. (2018). However, implementation of the foregoing strategies affects customer comfort and may not be consistent with the stochastic RE generation and customer demand patterns at all times. Hence, the case may lead to a high rate of dump energies and unmet demands. Moreover, energy curtailments may not be an optimal option as the process incurs additional losses and increased operational cost implications. Hence, optimal load management is appropriate for coordinating customer demands based on real-time supply availability. In terms of emission control, there may also be the need for intensive research in real estate to focus on the development of NZEBs, as demonstrated by Martirano et al. (2017) and Galisai et al. (2019).

Table 4 Critical analysis of DSM implementation methods in the literature
Full size table
Table 5 Demand side management strategies discussed in the literature
Full size table
Table 6 General concepts on tariff regimes discussed in the literature
Full size table

The factors of environment, economy and other social benefits continue to influence the proliferation of modern building system infrastructure as a global priority among most global societies. Hence, the importance of sustainability and energy conservation enhancement is emphasised among home users of energy. The sustainability and energy conservation emphasis is viewed here to influence electricity consumption patterns and affect customer comfort. It is worthy of note that the strategies used in energy management in buildings mainly focused on energy efficiency objectives. Hence, the BEMS is a strategic tool for intelligent grid management, as it allows interaction among building administrators, customers and utilities for monitoring, regulation and accountability in buildings’ energy utilisation (Sivaneasan et al. 2014).

The conventional time-of-use (\({C}_{TOU}\)) methods were considered in achieving load management schemes, as indicated by Pan et al. (2019), J. Liu and Zhong (2019), Rubaiee et al. (2019), Oprea et al. (2019) and Chen et al. (2019). Incidentally, the \({C}_{TOU}\) is fixed and does not usually appear to reflect the DSM-based desired operational objectives. It can be recalled that uncertain demand patterns hardly match RE resource distributions that periodically differ with the change in climatic conditions. This implies that the inflexibilities of \({C}_{TOU}\) may not be capable of enabling customer participation in optimal load management, where a close matching of stochastic RE generation and customer demands is the main goal. Methods such as real-time-pricing (RTP) can be considered in implementation of load management due to the customer-oriented features. Moreover, \({C}_{TOU}\) is utility-centred, and RTP emphasises on energy price per unit generation for a given time step. Hybridization of \({C}_{TOU}\) and RTP may enable customer’s full participation, through application of actual energy consumption charges. The proposed hybridization is envisaged to enable flexible options for customers’ decision on load scheduling. The benefits of load management expected to be achieved using the foregoing hybridization could be a multi-purpose strategy, with many DSM features such as the load shifting, peak shaving, energy arbitrage, strategic load growth and flexible load scheduling.

Conclusions

In renewable energy (RE) systems, optimal planning and operation schemes are implemented using optimization techniques. The methods ensure economic selection and placement of the RE components based on the location’s resource availability. The RE system operations cover the implementation of power dispatch and demand side management (DSM) schemes. The power dispatch is strategic to RE system operations. However, the RE resource intermittence and demand uncertainties render power dispatch complex and uneconomical. Unstable climatic conditions also contribute to uncertainties in the RE resources and customer demands. An instance is the electricity demand for space comforts and lighting in buildings that seasonally vary. Hence, the DSM is preferred, as it follows supplies. The DSM also ensures customer participation in the system’s operational management. In emerging smart energy management systems such as the nearly zero energy buildings, the conflicting objectives of economy, environment and customer comforts are efficiently reconciled. Hence, the DSM implementations through demand response (DR) programmes concerning energy cost and utility tariffs are suggested using any of the suitable DSM strategies discussed. This implies that the DSM strategies can be implemented by integration with the customer-oriented tariffs for residential and commercial demands such as the conventional time-of-use (\({C}_{TOU}\)), real-time pricing (RTP) or both.