Powering Future Advancements and Applications of Battery Energy Storage Systems Across Different Scales

Abstract

Battery Energy Storage Systems (BESSs) are critical in modernizing energy systems, addressing key challenges associated with the variability in renewable energy sources, and enhancing grid stability and resilience. This review explores the diverse applications of BESSs across different scales, from micro-scale appliance-level uses to large-scale utility and grid services, highlighting their adaptability and transformative potential. This study also includes advanced applications such as mobile energy storage, second-life battery utilization, and innovative models like Energy Storage as a Service (ESaaS) and energy storage sharing. Additionally, it discusses the integration of machine learning (ML) and large language models (LLMs), including advanced reinforcement learning (RL) algorithms, to optimize BESS operations and ensure safety through dynamic and data-driven decision-making. By examining current technologies, modeling methods, and future trends, this review provides a comprehensive overview of BESSs as a cornerstone technology for sustainable and efficient energy management, leading to a resilient energy future.

1. Introduction

In the evolving landscape of energy systems, Battery Energy Storage Systems (BESSs) stand at the forefront of technological innovation, offering a variety of solutions to some of the most pressing challenges in energy management and sustainability [1]. The latest BESS technologies, such as zinc-based batteries, offer promising pathways to address energy storage challenges, combining affordability, safety, and environmental sustainability [2,3,4,5,6]. As the demand for reliable, renewable, and resilient energy sources intensifies, BESS technology emerges as a pivotal element in the transition toward more sustainable and efficient energy systems [7].

The increasing reliance on renewable energy sources such as solar and wind is accompanied by inherent challenges related to their variability and intermittent nature. BESSs provide a crucial solution by enabling the storage of surplus energy during periods of high production and its subsequent discharge during periods of low production or high demand [8]. This capability not only helps stabilize the grid but also maximizes the utilization of renewable energy, making it a basis for the development of sustainability.

Moreover, BESS technologies contribute significantly to enhancing the resilience of energy systems [9]. In regions vulnerable to natural disasters or where energy infrastructure may be less reliable, BESSs can maintain power supply continuity during disruptions, thus safeguarding essential services and mitigating the impact on communities. The ability of these systems to quickly respond to fluctuations in power demand and supply further underscores their role in creating robust and responsive energy infrastructures.

However, the deployment of BESSs is not without challenges [10]. Economic factors, such as the high initial cost of installation and ongoing maintenance, pose significant barriers. Technological challenges also exist, including issues related to battery lifespan, efficiency, and the environmental impact of battery manufacture and disposal. Safety concerns, particularly regarding the risk of fires associated with certain types of battery chemistries, add another layer of complexity to the widespread adoption of BESSs. Despite these challenges, the opportunities presented by BESSs are substantial. Technological advancements continue to enhance the efficiency, capacity, and safety of these systems. Innovations in battery chemistry and management systems are expanding the potential applications of BESSs, from small-scale residential uses to large-scale industrial and utility implementations. As these technologies mature, they are expected to become more cost-effective and environmentally friendly, further accelerating the integration of BESSs into global energy markets.

This review aims to provide a comprehensive analysis of the applications and advancements of BESSs across various scales, from micro-scale appliance-level uses to large-scale utility applications. The manuscript specifically focuses on both traditional BESS technologies and their advanced application over the last five years to capture the most recent trends, innovations, and research progress in this rapidly evolving field. While several works have explored the applications and advancements of BESS, this review differentiates itself by focusing on emerging technologies, AI-driven optimization techniques, second-life battery applications, innovative service models like Energy Storage as a Service (ESaaS), energy sharing, etc., which have gained significant traction in recent years. These applications underscore the dynamic nature of BESSs as a solution not just for current energy demands but as a foundation for future energy landscapes.

By concentrating on the most recent developments, this review aims to provide insights into the evolving state of the field, highlighting advancements that were not extensively covered in previous reviews. This includes new approaches to thermal management, cost reduction strategies, predictive maintenance, and global regulatory frameworks for BESSs.

In our comprehensive review of 76 references related to BESSs, the majority of the studies—54 references (approximately 71%)—were published within the last five years (2020–2024). This significant concentration highlights the rapid advancements and heightened research interest in BESS technologies and applications in recent times. Additionally, 20 references (around 26%) fall within the 5–10-year range (2010–2019), representing foundational work that has established the current understanding and development trajectories in the field. Only two references (about 3%) are older than ten years (before 2010), providing essential historical context and early insights into BESS technologies. This distribution underscores the dynamic growth and evolving focus on BESS research, with contemporary studies driving innovation and addressing modern challenges in energy storage solutions.

2. Technology Overview

2.1. Basic Principles of BESSs

BESSs are integral components of modern energy management, bridging the gap between intermittent renewable energy production and the consistent demand for electricity [11]. BESSs are designed to store excess electrical energy produced during peak renewable generation times and then discharge that energy when demand is more than supply. This functionality not only stabilizes the grid but also maximizes the use of renewable sources by mitigating their inherent variability.

The operational principle of BESSs centers on the conversion of electrical energy into chemical energy and vice versa. This conversion is facilitated by the electrochemical reactions occurring within the structure of the BESS, which typically includes electrodes (anode and cathode), an electrolyte, and a separator [12]. The electrodes store the ions that are central to the battery’s energy storage capability. During the charging phase, electricity from external sources drives ions from the cathode to the anode, storing energy in the process. Conversely, during discharge, these ions move back from the anode to the cathode, releasing the stored energy as electricity. The electrolyte plays a crucial role by facilitating the movement of ions between the electrodes during the charge and discharge cycles. It is composed of salts, solvents, and additives that optimize ion flow, which is vital for maintaining the efficiency and lifespan of the battery. The separator, meanwhile, acts as a physical barrier between the anode and cathode, preventing direct contact that could lead to short circuits while still allowing for the passage of ions.

The efficiency of a BESS is often measured by its round-trip efficiency, which is the ratio of energy output during discharging to the energy input during charging. This efficiency is affected by the type of battery technology, component quality, and operational strategy.

Beyond these basic functions, BESS technology is evolving to meet increasing demands for higher energy density, faster charging times, and longer lifecycles. Innovations in battery materials and chemistry, such as the development of solid-state batteries and enhancements in lithium-ion technology, are at the forefront of this research. These advancements promise not only to improve the efficiency and safety of BESSs but also to extend their applications beyond simple energy storage. A crucial component in the operation of a BESS is the battery management system (BMS) [13], which ensures safe battery operation by monitoring their state, calculating secondary data, controlling the environment, and balancing the load among batteries. The BMS optimizes battery health and lifespan, ensuring that batteries operate within their safe operating area.

As BESS technologies evolve, they are becoming more sophisticated with greater capacities and efficiencies, which are essential for transitioning to an energy landscape dominated by renewable sources. Understanding these basic principles provides a foundation for integrating BESSs effectively into the energy grid, enhancing stability, efficiency, and sustainability.

2.2. Type of Battery

A BESS employs various battery types, each with unique characteristics and suitable applications. Understanding these differences is crucial for selecting the appropriate battery technology for specific energy storage needs. Here is a look at some of the most common types of batteries used in BESSs:

Lead–Acid Batteries: One of the oldest types of rechargeable batteries, lead–acid batteries are widely used due to their cost-effectiveness and reliable performance. They are typically used in backup power and load-leveling applications. However, they have a relatively low energy density and a shorter lifecycle compared to other types.

Lithium–Ion Batteries: Currently the most popular choice for BESSs, lithium-ion batteries offer high energy density and a longer lifespan. They are favored for their efficiency and versatility, finding applications across residential, commercial, and grid-scale storage. Their ability to quickly charge and discharge makes them ideal for applications requiring high-power bursts.

Nickel–Cadmium (NiCd) Batteries: Known for their robust performance in extreme conditions, NiCd batteries can endure a wide range of temperatures and have a long shelf life. They are often used in industrial applications where durability is crucial. However, their use has declined due to environmental concerns related to cadmium toxicity.

Nickel–Metal Hydride (NiMH) Batteries: NiMH batteries are considered a more environmentally friendly alternative to NiCd batteries, offering moderate energy density and fewer disposal concerns. They are commonly used in consumer electronics and hybrid electric vehicles but are less common in large-scale energy storage.

Flow Batteries: Unlike traditional batteries that store energy in solid electrodes, flow batteries store energy in liquid electrolytes, which are pumped through an electrochemical cell. This type allows for the decoupling of power and energy ratings, making them highly scalable and suitable for long-duration energy storage. Vanadium redox and zinc–bromine are two common types of flow batteries.

Sodium–Sulfur (NaS) Batteries: Operating at high temperatures and capable of storing large amounts of energy, NaS batteries are used primarily in grid storage applications. They have high energy density and efficiency but require significant safety measures due to their high operational temperatures and the corrosive nature of the materials involved.

2.3. Comparison of Performance Characteristics

The performance of BESSs is critically influenced by the characteristics of the batteries they employ. When comparing different types of batteries, several key performance metrics must be considered to determine their suitability for specific applications. Here is an overview of how these characteristics can vary among the most commonly used battery types:

Energy Density: This refers to the amount of energy a battery can store relative to its weight or volume. Lithium-ion batteries show advantages in this area, offering high energy density, which makes them ideal for applications where space and weight are limiting factors, such as in electric vehicles and portable electronic devices. Lead–acid batteries, on the other hand, have lower energy densities, making them bulkier and less suited for mobile applications.

Power Density: Unlike energy density, power density measures the speed at which a battery can release its stored energy. This is crucial for applications that require quick bursts of energy, such as for regulating grid frequency or providing emergency backup power. Lithium-ion batteries also perform well in terms of power density, providing rapid energy discharge when needed.

Efficiency: The round-trip efficiency of a battery measures the percentage of input energy that can be successfully converted back into usable electricity. Lithium-ion batteries typically have higher efficiencies, often around 90–95%, meaning less energy is lost in the charge/discharge cycle. Flow batteries also exhibit high efficiencies, particularly in scenarios where they are cycled frequently.

Lifespan and Lifecycle: The lifespan of a battery is determined by how many complete charge and discharge cycles it can undergo before its capacity falls to an unusable level. Nickel–metal hydride and lithium-ion batteries tend to have longer lifespans, with thousands of cycles possible under optimal conditions. Lead–acid batteries usually have shorter lifecycles and may degrade faster, especially if deeply cycled regularly.

Cost: The initial capital cost and the cost per cycle are important economic considerations. Lead–acid batteries generally have the lowest upfront costs, but their shorter lifespans and greater maintenance needs can increase their total cost of ownership. Lithium-ion batteries, while more expensive initially, may offer lower lifetime costs due to their longer operational life and lower maintenance requirements.

Safety and Environmental Impact: Safety concerns primarily revolve around the risk of fire and chemical leakage. Lithium-ion batteries, for instance, require sophisticated management systems to prevent overheating and potential fires. Regarding environmental impact, the toxicity of materials used (like cadmium in NiCd batteries or sulfur in NaS batteries) and the challenges associated with disposal or recycling are significant concerns that need to be addressed.

Scalability: This refers to how easily a technology can be scaled up or down to meet specific energy requirements. Flow batteries are highly scalable due to their unique design, allowing for increased energy storage capacity simply by enlarging the storage tanks of electrolytes.

Figure 1 shows an intuitive comparison between them [14,15,16,17,18,19,20,21,22,23].

2.4. Mathematical Modeling

Mathematical modeling of BESSs plays a critical role in predicting system performance, optimizing design, and ensuring efficient operation and management. This section outlines the fundamental mathematical approaches used to model different aspects of BESSs.

2.4.1. SOC Modeling

The State of Charge (SoC) of a battery is a key parameter that indicates the remaining capacity of the battery as a percentage of its total capacity, shown in Equations (1) and (2). Accurately modeling SoC is vital for effective battery management and longevity.

$$SO{C}_t=E_t^{BESS}/E^{BESS,\max }$$

(1)

$$SO{C}_t=SO{C}_{t-1}+{\displaystyle \frac{P_t^{BESS,cha}\Delta t{\eta }^{cha}}{E^{BESS,\max }}}-{\displaystyle \frac{P_t^{BESS,dis}\Delta t}{E^{BESS,\max }{\eta }^{dis}}}$$

(2)

where $SO{C}_t$ is the SOC of the BESS at time t; $E_t^{BESS}$ is the energy stored in the BESS; $E^{BESS,\max }$ is the maximum capacity of the BESS; $P_t^{BESS,cha}$ is the charging power; $P_t^{BESS,dis}$ is the discharging power; ${\eta }^{cha}$ is the charging efficiency; ${\eta }^{dis}$ is the discharging efficiency.

2.4.2. DoD and Battery Degradation Modeling

Depth of discharge (DoD) is a crucial concept in battery technology, particularly when discussing battery life, performance, and capacity in BESSs [24]. DOD refers to the amount of energy that has been discharged from a fully charged battery, expressed as a percentage of the total capacity of the battery. Essentially, it measures how deeply a battery is discharged before it is recharged.

Battery degradation is a complex process influenced by several factors, including cycling, temperature, depth of discharge, rate of charge or discharge, and the intrinsic chemistry of the battery itself. To accurately predict the lifespan and performance decline of batteries in BESSs, several mathematical models have been developed. A commonly used model is presented to understand and predict battery degradation. The semi-empirical Capacity Fade Model combines empirical data with basic electrochemical principles to predict the rate of capacity fade over time and usage. The capacity fade of the battery can be modeled as a function of the cycle number and DoD.

$$E_{fade}^{BESS,\max }=E_{initial}^{BESS,\max }\exp \left(-{\left({\displaystyle \frac{n}{A}}\right)}^{\alpha }-{\left({\displaystyle \frac{DoD}{B}}\right)}^{\beta }\right)$$

(3)

where $E_{initial}^{BESS,\max }$ is the initial capacity of the BESS; $n$ is the number of cycles used; $A$, $B$, $\alpha$, and $\beta$ are empirical constants specific to the battery chemistry and configuration.

2.4.3. SoH Modeling

State of Health (SoH) is a crucial metric in battery management systems, indicating the overall condition and health of a battery relative to its ideal conditions when new. SoH can be influenced by various factors, including the number of charge/discharge cycles, DoD, operating temperatures, and the rate at which the battery is charged or discharged. Below is a common approach to modeling SoH, focusing on capacity fade and internal resistance increase, two primary indicators of battery health degradation. Capacity fade over time is one of the primary indicators of battery SoH. It can be modeled using an exponential decay function to reflect the loss of capacity due to cycling and aging:

$$SoH=E^{BESS,\max }/E_{initial}^{BESS,\max }\times 100\%$$

(4)

Another approach is using cycle-based degradation models that focus on the accumulation of damage with each charge–discharge cycle, considering factors such as temperature and charge rate.

$$\Delta SoH=k\cdot f\left(DoD,T,{\eta }^{cha/dis}\right)$$

(5)

where $k$ is the degradation rate constant; $f\left(DoD,T,{\eta }^{cha/dis}\right)$ represents the combined effects of DoD, temperature, and the charge/discharge rate.

2.4.4. Lifespan Modeling

The lifespan of a battery in a BESS is influenced by several factors, such as the number of charge–discharge cycles, DoD, operating temperatures, charging practices, and the inherent stability of the battery chemistry [25]. Here is a basic approach to modeling the lifespan of batteries using the lifecycle as the primary metric:

$$L{f}_s={\displaystyle \frac{N_{100}^{end}}{365\cdot N_{100}^{eq,day}}}={\displaystyle \frac{N_{100}^{end}}{365\cdot {\displaystyle \sum _t{\left(\Delta Do{D}_{t,s}\right)}^{kp}}}}$$

(6)

where $N_{100}^{end}$ is the lifecycle of the BESS under a DoD of 100%; $N_{100}^{eq,day}$ is the equivalent number of the cycle under a DoD of 100%; $\Delta Do{D}_{t,s}$ is the increase in the battery DoD during continuous discharging; $kp$ is the Peukert lifetime constant, representing the dependence of the lifecycle on DoD (ranging from 0.8 to 2.1).

3. Scaling Success: BESS Deployment from Micro to Macro Across Complex Energy Networks

The deployment of BESSs across different scales—from micro to macro—plays a crucial role in transforming energy networks into more complex, integrated, and responsive systems. The BESS at different levels plays a vital role across energy networks by managing demand peaks, enhancing renewable energy integration, providing reliable emergency backup, and supporting grid services. This multifunctional capability makes BESSs essential for advancing a resilient, efficient, and sustainable energy infrastructure. Figure 2 shows an overview of BESS applications across various scales.

3.1. Micro-Scale: Appliance-Level Applications

At the micro-scale, a BESS is integral to powering a wide range of individual devices where portability, autonomy from the electrical grid, and precise energy management are paramount [26]. This application level typically includes various consumer electronics, medical equipment, and portable power tools. Here, the focus is on optimizing the size and efficiency of the storage solutions to enhance the overall utility and performance of small-scale devices.

In the realm of consumer electronics, such as smartphones, laptops, tablets, wireless headphones, and digital cameras, micro-scale BESSs are crucial. They not only provide the necessary power to operate these devices but also influence design innovations, allowing for thinner, lighter, and more energy-efficient models. The continuous advancement in battery technologies, particularly with lithium-ion and emerging solid-state batteries, has significantly extended the operational time of these devices between charges, enhancing user convenience and efficiency.

Portable power tools and other handheld devices also benefit greatly from developments in micro-scale BESSs. These tools rely on compact yet powerful batteries to offer freedom from wired power sources without sacrificing performance. As such, BESSs enables higher productivity and greater accessibility in various applications, from construction sites to remote field services.

3.2. Small Scale: Residential Applications

At the small scale, BESSs are revolutionizing residential energy management by enabling homeowners to optimize their energy usage, decrease reliance on the grid, and enhance sustainability [27]. These systems are often paired with renewable energy sources such as solar panels, creating integrated, efficient, and self-sustaining home energy solutions.

One of the primary applications of small-scale BESSs in residential settings is for solar energy storage. Homeowners with solar installations can use BESSs to store excess energy generated during peak sunlight hours in plug-and-play mode [28]. This energy is then available for use during the evening or on cloudy days, effectively reducing dependence on grid-supplied electricity and lowering electricity bills. The ability to store and use solar energy on demand maximizes the utilization of installed solar panels, providing significant cost savings over time.

Additionally, in regions vulnerable to power outages due to weather disturbances or grid failures, small-scale BESSs serve as an important source of backup power [29]. These systems ensure a continuous power supply, safeguarding homes against interruptions and providing peace of mind. The ability to maintain power during outages is particularly crucial for safety and convenience, as it keeps critical systems such as lighting, heating, and refrigeration operational.

In addition to the traditional uses of small-scale BESSs in residential settings, the integration of electric vehicles (EVs) and Vehicle-to-Everything (V2X) technology represents a dynamic and rapidly evolving dimension of home energy management [30]. EVs can be considered mobile BESS units, storing electricity that can not only power the vehicle but also be utilized for home energy needs. The batteries in EVs are typically larger and more capable than those used solely for home energy storage, providing a significant amount of energy that can be tapped into as needed. As EV adoption increases, the potential for these vehicles to contribute to residential energy management grows exponentially.

V2X technology extends the functionality of EVs by enabling them to interact with the home and the wider electricity grid in multifaceted ways. One of the most significant applications of V2X is Vehicle-to-Home (V2H), where the energy stored in an EV’s battery is used to power the home during peak times or outages [31]. This not only provides a backup power source but also assists in peak shaving, similar to stationary home battery systems, but with the added advantage of mobility. V2X also includes Vehicle-to-Grid (V2G) interactions, where EVs can feed energy back into the grid, supporting grid stability and allowing homeowners to earn rebates or credits from utility companies [32]. This two-way interaction helps balance demand and supply on the grid, making renewable energy usage more efficient and reducing overall carbon emissions.

3.3. Medium Scale: Commercial, Industrial, and Community-Level Applications

At the medium scale, BESSs play a pivotal role in enhancing energy management for commercial establishments, industrial operations, and community infrastructures. This segment of BESS applications focuses on optimizing energy usage, reducing operational costs, and improving reliability and sustainability across broader areas than individual residences.

In commercial and industry settings such as offices, retail stores, manufacturing, and small businesses, BESSs are utilized to manage energy costs effectively by engaging in peak shaving and demand charge reduction [33]. Similar to the mode at the residential scale, medium-scale BESSs store energy during off-peak periods when it is cheaper and discharges it during peak demand times at a large scale. Owners can significantly reduce the high costs associated with peak energy rates. Additionally, BESSs can provide emergency backup power, which is crucial for maintaining operations during power outages, thus ensuring business continuity and protecting critical data and systems. Furthermore, BESSs help in managing expenses and can even participate in energy arbitrage opportunities.

Virtual Power Plants (VPPs) represent an innovative medium-scale application of BESSs, offering a dynamic approach to managing and utilizing distributed energy resources (DERs) across various locations [34]. VPPs employ advanced software and control systems to aggregate the capacities of numerous small-scale power-generating units and energy storage systems, essentially functioning as a single flexible and responsive power plant [35,36]. The core idea behind a VPP is to leverage the collective power of decentralized energy resources such as commercial and community BESSs [37]. These resources are connected and managed through sophisticated software that optimizes power production and distribution based on real-time demand and market conditions. By coordinating the operation of these disparate units, a VPP can either supply energy to the grid or absorb excess energy, just as a traditional power plant would, but with enhanced efficiency and agility.

3.4. Large Scale: Utility-Scale Applications and Grid Services

At the large scale, BESSs are also critical components in utility-scale applications and grid services. These systems are deployed to improve grid reliability, enhance operational flexibility, and facilitate the integration of renewable energy sources [38]. Large-scale BESSs are pivotal in addressing the challenges posed by the increasing complexity and demands of modern electrical grids.

Utility-scale BESSs are typically large installations that store substantial amounts of energy, which can be dispatched when needed to meet consumer demand or maintain grid stability. These systems are often connected directly to the grid at substations or near renewable generation sites like wind farms or solar parks.

Beyond energy storage, large-scale BESSs provide a range of vital grid services. These include frequency regulation, where the BESS quickly responds to changes in grid frequency to correct imbalances between supply and demand, and voltage support, which is crucial in areas with high levels of renewable generation that can cause voltage fluctuations [25]. Additionally, BESSs can be used for peak shaving, reducing the load on the grid during times of high demand and alleviating stress on grid infrastructure.

BESSs also play a critical role in enhancing grid resilience, providing black start capabilities that are essential for restoring power efficiently after significant outages [39]. This function allows BESSs to supply the necessary energy to start up the grid’s main generators, a process critical in minimizing downtime during blackouts.

Looking forward, the importance of large-scale BESSs in utility applications and grid services is expected to increase significantly. As grids become more digitized and incorporate higher levels of renewable energy, the flexibility and rapid response capabilities offered by BESSs will become increasingly crucial. Coupled with advancements in battery technology and decreasing storage costs, these factors are set to expand the role of BESSs, making them fundamental components in achieving more sustainable, reliable, and efficient power systems [40].

4. Proactive Applications: Pioneering the Future of Battery Energy Storage

4.1. On the Move: Exploring Mobile Energy Storage Solutions

4.1.1. Application Scenarios

In recent years, with the process of transportation electrification, the concept of mobile power sources has emerged. Compared with stationary energy storage systems (SESSs), the mobility of mobile power sources enhances their capability of tapping into multiple value streams that have spatiotemporal variability, which in turn improves their asset utilization and potentially their value proposition [41]. Mobile energy storage systems are crucial for applications ranging from EVs to utility-scale storage banks carried by trucks that require high-density energy storage for extended mobility. The mobility of these storage solutions enables a swift and flexible response to varying energy demands, facilitating energy distribution in disaster-struck areas or regions undergoing temporary energy shortages. Moreover, mobile BESSs can be integrated into microgrid configurations, enhancing grid resilience and stability by providing supplementary power and balancing services as needed [42].

Typically, MESS units are charged at a central depot and then dispatched to established base stations where they can supply power to the grid or to critical users, such as hospitals or research laboratories, during emergencies. The basic conceptual framework of MESS can be shown in Figure 3. This dynamic deployment strategy not only maximizes the utility of storage systems but also plays a crucial role in maintaining energy reliability in critical situations. The concept of Uber Energy Storage Systems (UESSs) offers a distinct advantage by emphasizing on-demand service provision [43]. This approach connects service providers and receivers in real time through advanced communication infrastructure. In UESSs, consumer requests and the state information of the energy storage systems are continuously collected and processed, requiring a centralized platform that can efficiently match service orders with available UESS units to ensure timely and responsive services.

4.1.2. Modeling Methods

The power dispatch model of MESSs should be designed to be seamlessly integrated with conventional power system components, including generators, SESSs, and other grid infrastructure. This integration is crucial for ensuring that MESSs can effectively complement existing energy resources, providing additional flexibility and resilience to the power grid [44]. Unlike an SESS, which is fixed and primarily interacts with the local grid, an MESS is inherently mobile and dynamic. This mobility allows MESSs to serve as moving energy resources that can deliver power precisely where and when it is needed within the transportation network. Hence, the modeling of MESSs requires transportation network modeling [45]. Models need to account for varying traffic patterns and mobility behaviors to predict where MESSs will be most effective. This involves using traffic flow simulations and mobility data to analyze patterns that influence energy consumption and requirements. Incorporating temporal–spatial dynamics into the modeling of systems such as MESSs or transportation networks is essential for capturing the variations in both time and space that characterize their operations [46]. This approach helps in accurately predicting how these systems will behave under different conditions and in different locations, providing insights that are crucial for optimization and strategic planning.

Comparing the models of MESSs and UESSs, the operation strategy for MESSs is typically optimized on a day-ahead basis, making it more rigid and heavily dependent on accurate predictions of future states. This reliance on forecasts can limit the flexibility and responsiveness of MESSs in dynamic environments. However, a UESS’s real-time order dispatch mechanism significantly enhances the flexibility, operational efficiency, and commercialization potential of mobile energy storage services. Therefore, it requires a more rapid and efficient dispatch algorithm [43]. This allows a UESS to quickly respond to unforeseen circumstances or emergencies, making it a more agile and effective solution in rapidly changing or unpredictable scenarios.

When designing and implementing an on-demand MESS service that interacts with transportation networks, considering the delay factors associated with transportation and navigation is crucial. These factors directly impact the efficiency, reliability, and overall effectiveness of the service [46]. Traffic congestion is a primary cause of delays in urban and suburban areas. It can significantly affect the travel time of MESS vehicles, especially during peak hours, reducing the system’s ability to respond promptly to energy demands. Efficient routing is critical for minimizing delays. Navigation systems must be optimized to choose routes that avoid known congested areas or roadworks, adapting in real-time to traffic updates and changes in road conditions. To solve this problem, it also requires detailed modeling of integrated electricity and transportation networks based on real-time road monitoring and grid condition predictions. Additionally, strategically positioning MESS resources in high-demand areas based on predictive models can reduce the need for long-distance travel, thereby minimizing the impact of potential delays. This approach requires a good understanding of typical energy usage patterns in different areas.

4.2. Renewed Purpose: Second-Life Applications of BESSs

4.2.1. Application Scenarios

As the adoption of BESSs in EVs and other applications grows, the topic of what to do with batteries once they reach the end of their primary lifecycle becomes increasingly relevant. These batteries, often still capable of significant storage capacity despite no longer meeting the performance requirements for their original use, present an opportunity for second-life applications [47]. This approach not only extends the useful life of the batteries but also contributes to environmental sustainability by reducing waste and the demand for raw materials.

A second-life BESS refers to the practice of reusing batteries from electric vehicles and other applications in new energy storage roles after they are no longer fit for their initial purpose; the process is shown in Figure 4. Typically, an EV battery is considered to have reached the end of its first life when its capacity falls below 80% of its original state, which may reduce its effectiveness for vehicular use but still allows for substantial storage capacity for less demanding applications. Reusing batteries reduces the cost barrier to deploying energy storage systems, making them more accessible for various applications [48]. Furthermore, it decreases the environmental impact by extending the life of the batteries and reducing the need for new materials through recycling and repurposing. Second-life battery projects can be scaled based on the available number of used batteries, allowing for flexible and adaptive energy storage solutions.

Batteries are typically retired from EVs when they can no longer sustain the high power demands required for automotive performance, generally when their capacity falls to about 70–80% of their original. However, these batteries still possess enough capacity and stability to be highly effective in residential settings [49]. The performance requirements for home energy storage are considerably less stringent than for vehicles. Residential applications generally require lower power outputs and endure longer, more consistent discharge cycles, making second-life batteries ideal for such uses.

At a larger scale, these batteries can still effectively serve grid services because the operational demands for these applications are significantly lower and less frequent compared to automotive uses [50]. Grid services such as load balancing, frequency regulation, and demand response typically involve slower and more predictable power discharges. These tasks do not necessitate the same level of instantaneous high energy output that driving does, making second-life batteries perfectly adequate for such roles.

The market for second-life BESSs grows as industries and governments increasingly focus on sustainability and circular economy principles. Innovations in battery health assessment, combined with advancements in system integration technologies, will likely drive the expansion of this market, making second-life applications a standard practice in the lifecycle of batteries.

For second-life BESSs, especially for large-scale deployment, safety is a concern due to the degradation of batteries previously used in EVs. These batteries may have reduced performance and higher failure rates. Thorough testing, refurbishment, and advanced BMSs are necessary to ensure safe operation. Additionally, second-life batteries require careful monitoring for issues like voltage imbalances or internal short circuits.

4.2.2. Modeling Methods

The initial step in modeling involves accurately assessing the remaining capacity and overall health of used batteries. This requires detailed testing to determine parameters like remaining energy capacity, power capability, and degradation rate. Techniques such as electrochemical impedance spectroscopy (EIS) and charge–discharge cycling are commonly used [51]. The results inform the potential suitability and lifespan of the battery in its second-life application.

Understanding how the battery will continue to degrade under its new usage conditions is crucial. This involves creating models that predict future capacity and performance based on historical degradation patterns, operational conditions, and expected usage in the second-life setting. Common models include empirical, semi-empirical, and physics-based models that consider factors like DoD, temperature, and charging patterns.

Assessing the economic feasibility of using second-life BESSs is vital [52]. This includes calculating the costs associated with the testing, reconditioning, transportation, and integration of the batteries into new systems. Additionally, models must consider potential revenue streams from applications such as demand response, load shifting, or grid services. Cost–benefit analysis and return on investment (ROI) calculations are integral to this aspect of modeling.

Safety is a primary concern when repurposing batteries, as aged batteries may pose increased risks of failure. Reliability models are used to predict the probability of failure modes such as thermal runaway or cell imbalance. This involves stress testing under simulated operating conditions to identify potential safety risks and implementing strategies such as enhanced BMSs for monitoring and mitigation.

Modeling also includes the design and integration of second-life batteries into new systems. This requires simulations to determine how batteries with varied histories and capacities can be optimally configured and managed. Issues such as battery mismatch, SoC balancing, and energy management strategies are key focuses in these models.

4.3. Energy Storage as a Service (ESaaS) and Energy Storage Sharing/Cloud Energy Storage

4.3.1. Application Scenarios

ESaaS is an innovative business model that is transforming how energy storage capabilities are deployed and utilized. These concepts allow businesses, utilities, and consumers to benefit from advanced energy storage solutions without the upfront investment and maintenance responsibilities typically associated with owning and operating battery storage systems. In ESaaS, there are typically multiple parties, such as battery owners, users (e.g., businesses, residential consumers), energy providers, and possibly third-party service providers. These parties may have conflicting goals, such as energy optimization versus cost reduction, or maximizing return on investment for battery owners versus providing affordable access to storage for consumers. Aligning these diverse interests requires careful structuring of agreements, clear definitions of roles and responsibilities, and the establishment of frameworks that ensure fair access and benefits for all parties.

ESaaS is a service model where customers pay for energy storage solutions as an ongoing service rather than an upfront product purchase [53]. This approach removes the capital expenditure (CAPEX) barrier, making advanced battery technologies accessible to a broader range of users. It typically includes the installation, operation, and maintenance of energy storage systems by a service provider, who also manages the performance and risk associated with the system.

Energy storage sharing is a typical type of ESaaS which involves multiple parties or customers sharing a communal energy storage system, as shown in Figure 5. This model is particularly effective in settings like apartment buildings, cooperative housing projects, or among small businesses in close proximity [54]. The shared system can be managed by a single entity or a collaborative group, spreading the cost and benefits across all participants [55]. Sharing a larger system can reduce individual user costs through economies of scale in both purchase and maintenance. Furthermore, a shared system can optimize energy usage across multiple points, improving overall efficiency and reducing wastage. In communal settings, shared energy storage can enhance resilience against power outages, providing a reliable backup for essential services [56]. Furthermore, profit-sharing mechanisms in energy storage-sharing models are another significant challenge. For example, when storage systems are used for services like peak shaving, grid stabilization, or participation in energy arbitrage, determining how the profits or savings are distributed among stakeholders can be complex. Battery owners may expect a return on their investment, while users might expect reduced energy costs, and energy providers could seek to maximize grid reliability and reduce operating costs. Developing transparent and equitable profit-sharing models that fairly distribute the financial benefits while incentivizing efficient system usage will be key to the success of ESaaS platforms.

Cloud energy storage, as another type of ESaaS, is similar to traditional shared energy storage but offers a more flexible and scalable approach by utilizing virtualized storage resources managed through a centralized digital platform. While energy storage sharing typically requires users to be in close physical proximity to access a communal storage system, cloud energy storage allows users to tap into distributed storage resources without geographical constraints. Users access and manage energy storage resources via a cloud-based platform. Unlike shared energy storage, cloud energy storage does not require users to be physically connected to the same energy storage system. This approach enables dynamic allocation of storage capacity based on real-time needs, making it particularly advantageous for businesses and utilities that operate across multiple locations or need to scale their storage solutions quickly.

In this energy storage sharing mechanism, prosumers can discharge energy into the shared BESS, which is effectively considered as borrowing capacity from the system managed by the coordinator. When the prosumer later charges energy from the shared BESS, it is viewed as returning capacity to the shared system [57]. To facilitate this exchange, a credit system is utilized to track the amount of capacity each prosumer has borrowed. Combined with blockchain technologies, this credit system is crucial for accurately accounting for the sharing fees, ensuring that each participant’s usage and contributions are fairly recorded and compensated within the shared energy storage framework.

4.3.2. Modeling Methods

One of the core methodologies in energy storage-sharing systems is game-theoretic modeling [58], which provides a structured approach to managing the strategic interactions between participants. These models treat each participant, whether a grid operator, prosumer (who both consumes and produces energy), or consumer, as a player with specific objectives and strategies. The primary goal of game-theoretic models is to identify Nash Equilibria, where no participant can improve their outcome by unilaterally changing their strategy. This approach is particularly beneficial in scenarios where participants must balance their individual goals with the collective objective of maintaining grid stability and efficiency. By considering the interactions and potential conflicts between different stakeholders, game-theoretic models help to optimize decision-making processes, ensuring that the overall system operates more efficiently [59]. By promoting cooperation and strategic alignment among participants, game-theoretic approaches contribute to a more resilient and well-coordinated energy storage system.

In scenarios where the allocation of energy storage resources needs to be managed in real time, auction-based mechanisms are highly effective [60]. These mechanisms typically employ a uniform-price or discriminatory-price auction format, where participants submit bids indicating the price they are willing to pay for a certain amount of storage capacity. The auction algorithm then determines the clearing price based on the intersection of supply and demand curves, ensuring that storage resources are allocated to those who value them the most. Advanced implementations of auction-based models might incorporate multi-round bidding processes, where participants can adjust their bids based on market signals, or combinatorial auctions, where participants bid on bundles of storage services that best meet their needs [61]. These mechanisms are particularly useful in managing limited storage capacity in a way that maximizes economic efficiency and ensures transparency in resource allocation.

The fair distribution of profits generated from energy storage sharing is critical for sustaining long-term cooperation among participants. Profit allocation methods such as the Nucleolus, Shapley value, and Nash bargaining solutions provide robust frameworks for equitable profit distribution [62]. The Nucleolus method seeks to minimize the maximum dissatisfaction among participants by iteratively adjusting profit shares until the smallest excess (or disagreement) is achieved across all coalitions. The Shapley value, derived from cooperative game theory, allocates profits based on the marginal contribution of each participant to the total value generated by the coalition, ensuring that participants are rewarded proportionately for their input. Nash bargaining solutions focus on achieving a mutually agreeable distribution of profits by maximizing the product of the participants’ utilities, taking into account their disagreement points. These methods are particularly important in multi-stakeholder environments, where ensuring fairness in profit distribution is key to maintaining a stable and cooperative energy-sharing ecosystem.

4.4. ML and AI in BESSs

4.4.1. Application Scenarios

ML and AI are increasingly becoming integral to optimizing the performance and enhancing the capabilities of BESSs. These technologies offer sophisticated analytical tools and adaptive learning capabilities that significantly improve decision-making processes, operational efficiency, and predictive maintenance strategies in energy storage applications. For instance, AI-driven predictive maintenance helps identify potential failures by analyzing historical data, reducing downtime, and extending system life. Energy dispatch optimization uses reinforcement learning to determine optimal charge/discharge schedules based on real-time factors like grid demand and market pricing, improving cost-effectiveness. SoC forecasting with deep learning ensures efficient battery operation by accurately predicting charge levels, while grid frequency regulation models adjust BESS operations to stabilize the grid. Additionally, AI enhances load forecasting and demand response, enabling peak shaving and efficient energy storage. Energy arbitrage optimization analyzes market price fluctuations to maximize economic returns by timing energy storage and release. These techniques collectively improve BESS efficiency and contribute to more resilient and cost-effective energy systems.

ML algorithms are exceptionally well suited for analyzing large datasets to predict battery degradation and other performance-related issues. By processing real-time and historical data on battery usage, charge cycles, temperature, and other critical parameters, ML models can forecast the future health and lifespan of the battery systems [63]. This predictive capability allows for proactive maintenance, preventing failures before they occur and extending the overall lifespan of the storage system.

Integrating BESSs with renewable energy sources is often challenging due to the variability and unpredictability of solar and wind energy production. AI and ML enhance this integration by accurately predicting renewable output and aligning it with storage capabilities and grid demand [64]. This synchronization ensures that the maximum amount of renewable energy is captured, stored, and utilized effectively, minimizing wastage and improving the stability of the grid. Furthermore, machine learning can also predict the SOH of the BESS according to its features [65,66].

Moreover, deep reinforcement learning (DRL) offers a sophisticated approach to optimizing BESSs by dynamically adjusting charging and discharging strategies [67]. Based on the framework of Markov decision processes (MDPs), DRL algorithms take into account a variety of factors, such as forecasts of energy demand, fluctuations in electricity prices, and patterns of renewable energy generation [68]. This complex decision-making process allows for real-time optimization that aligns with current market and environmental conditions. Utilizing DRL not only maximizes economic returns from energy arbitrage opportunities but also helps in reducing wear and tear on the batteries, thereby extending their operational lifespan. Additionally, the efficiency of energy usage is significantly enhanced, promoting a more sustainable and cost-effective management of energy resources.

Figure 6 demonstrates a neural network-based framework for AI applications in BESSs. The key features of BESSs are input into the neural network, which processes data through multiple hidden layers to generate the target outputs, such as SOH analysis or control actions in the DRL context. In the DRL context, these control actions optimize energy dispatch, charge/discharge strategies, and maintenance scheduling based on observations and rewards received from the environment. The environment includes grid conditions, electricity markets, and other external factors, highlighting AI’s role in dynamically improving BESS performance and efficiency.

4.4.2. Modeling Methods

Traditional ML algorithms, such as neural networks, particularly deep neural networks (DNNs), are highly effective in capturing complex nonlinear relationships within large datasets. For battery health prediction, DNNs can be trained on historical data encompassing charge cycles, load demands, temperature variations, and other operational metrics [69]. The networks learn to identify patterns and anomalies that indicate degradation, providing reliable predictions of the battery’s future performance and lifespan. This approach is especially powerful due to its ability to adjust to new data, continually refining its predictions as more information becomes available. Although primarily known for image processing, Convolutional Neural Networks (CNNs) have been adapted for sequential data analysis and are useful in analyzing time-series data from batteries [70]. They can extract features from cycle-by-cycle operational data, helping to predict how different usage patterns affect battery health and longevity. CNNs are particularly adept at handling spatial–temporal data, making them suitable for tasks where time-related patterns need to be extracted from multivariate time series.

Recurrent Neural Networks (RNNs) and their more sophisticated variant, Long Short-Term Memory Networks (LSTMs), are ideal for making predictions based on sequential data [71]. In the context of BESSs, they can analyze historical performance data over time to predict future degradation and SoH. LSTMs are particularly beneficial due to their ability to remember long-term dependencies, avoiding the vanishing gradient problem common in standard RNNs. This makes them exceptionally well suited for applications where past events significantly influence future outcomes, such as in cyclic aging processes in batteries.

Large language models (LLMs), particularly those based on Transformer architectures, can be innovatively adapted for predictive analytics in BESSs [72]. These models leverage advanced self-attention mechanisms, which are critical in processing and analyzing temporal sequences in energy usage, generation, and consumption data. The self-attention mechanism allows LLMs to focus on relevant parts of the input data at different time steps, capturing intricate temporal dependencies and relationships that are often missed by traditional time-series models. One of the key strengths of LLMs in this context is their ability to handle large, complex datasets and extract meaningful insights across multiple scales and dimensions. For example, LLMs can be trained on diverse datasets that include historical energy consumption patterns, weather forecasts, market prices, and operational constraints, allowing them to predict battery degradation, optimize load management, and forecast energy demand with high accuracy. By recognizing subtle patterns in the data, LLMs can also anticipate anomalies or shifts in energy usage, enabling proactive adjustments to BESS operations. Moreover, LLMs excel in multitask learning, where a single model can perform multiple predictive tasks simultaneously, such as predicting battery lifespan, forecasting energy prices, and optimizing charging schedules. This multitasking capability enhances the operational efficiency of BESSs by integrating various predictive functions into a cohesive framework. The adaptability of LLMs also makes them suitable for real-time applications. As new data become available, these models can continuously update their predictions, providing dynamic and up-to-date insights that support real-time decision-making. This is particularly valuable in managing the integration of renewable energy sources, where supply variability can be high, and accurate, real-time predictions are essential for maintaining grid stability. In addition, LLMs can be fine-tuned for specific contexts within the BESS framework, such as optimizing energy storage in response to specific regulatory environments or market conditions. This fine-tuning capability ensures that the models remain relevant and effective in various operational settings, from residential storage systems to large-scale utility deployments.

RL can optimize decision-making processes concerning battery usage and maintenance strategies [73]. By defining the problem as a Markov decision process, RL algorithms learn the best policies for charging and discharging, considering the long-term health and efficiency of the battery. This method is proactive, focusing on maximizing the lifespan and utility of the battery system by learning from each interaction with the system to continuously improve operational decisions. Algorithms such as Deep Q-Networks (DQNs) and their variants (Double DQN and Dueling DQN) offer robust solutions for discrete action spaces, making them suitable for decisions like when to charge or discharge batteries based on price signals and demand forecasts [74]. Proximal Policy Optimization (PPO) and Trust Region Policy Optimization (TRPO) are favored for their stability and effectiveness in environments with continuous action spaces, ideal for fine-tuning power outputs or managing real-time energy distribution [75]. Some actor-critic mechanism-based algorithms, such as Soft Actor Critic (SAC) and Twin Delayed Deep Deterministic Policy Gradient (TD3), significantly enhance the management of BESSs [76]. SAC, known for its sample efficiency and stability, optimizes a stochastic policy in an off-policy manner and incorporates an entropy bonus to encourage exploration, making it highly effective for dynamic and uncertain BESS operations. In addition, TD3 addresses the overestimation issues common in actor-critic methods by maintaining two separate Q-functions and using the smaller of the two for policy updates, which enhances the robustness and stability of the learning process. This method is particularly well suited for the continuous action spaces in BESSs, such as dynamically adjusting charging rates or discharging schedules based on real-time grid demands and storage conditions. To further consider the safety constraints of BESSs, some safe DRL algorithms have been designed to incorporate safety constraints directly into the decision-making process. These algorithms, such as Safe Policy Optimization (SPO) and Constrained Policy Optimization (CPO), are designed to balance the exploration–exploitation trade-off while adhering to predefined safety thresholds, such as avoiding overcharging, preventing thermal runaway, or minimizing degradation rates.

5. Challenges

The widespread adoption of innovative BESS applications faces several regulatory and operational challenges that could hinder their effectiveness and scalability. One significant challenge is the lack of standardized regulations across different regions and markets. With diverse energy policies and regulatory frameworks, there is a lack of consistency in how BESS technologies are integrated into the grid. This disparity complicates the development and deployment of BESS solutions across borders, making it difficult for manufacturers and operators to adhere to multiple, often conflicting, standards. For instance, battery safety standards, performance testing protocols, and environmental regulations may vary, leading to increased costs and delays in product development and deployment.

Another challenge is the interoperability of BESSs with existing grid infrastructure. As BESSs become more complex with the integration of AI and machine learning for optimized operations, ensuring seamless communication between various systems and devices becomes crucial. The absence of universal communication protocols or a unified interface between different BESS components and grid management systems could lead to inefficiencies and operational issues.

Grid integration is another hurdle, as current grid infrastructure may not be designed to handle large-scale energy storage. Many power grids, especially in developing regions, may lack the necessary flexibility and technological capacity to support widespread BESS integration without significant upgrades. Regulatory bodies may also face difficulties in creating frameworks that allow for the smooth operation of BESSs in dynamic markets while ensuring grid stability.

Additionally, the data privacy and security concerns surrounding AI-driven BESS applications pose another challenge. As BESSs collect and process large amounts of data, including real-time energy consumption and behavioral patterns, ensuring the security and confidentiality of these data becomes critical. Without clear regulations on data privacy, stakeholders, particularly consumers, may be reluctant to adopt BESS technologies.

Lastly, there are economic barriers, such as the high initial capital investment required for BESS infrastructure. While the costs of batteries have been decreasing, the upfront costs for large-scale deployment of BESSs remain significant. Without strong financial incentives or subsidies, many organizations may be reluctant to invest in BESSs, particularly in regions where energy storage is not yet economically viable.

6. Conclusions and Recommendations

BESSs are rapidly becoming a critical component of modern energy networks, offering versatile solutions across multiple scales, from individual appliances to large-scale utility applications. This review highlights the transformative potential of BESSs in enhancing grid stability, integrating renewable energy, and providing essential services like load balancing, frequency regulation, and backup power. Advanced applications such as second-life battery utilization, mobile energy storage, and innovative service models like ESaaS further expand the scope and impact of BESSs. The integration of ML and advanced deep reinforcement learning algorithms offers promising pathways to optimize BESS operations, ensuring not only efficiency but also safety and reliability in energy management.

As the technology continues to evolve, addressing challenges such as cost, degradation, and standardization will be critical to maximizing the benefits of BESSs. By leveraging advanced modeling techniques and AI-driven approaches, the deployment of BESSs can be significantly enhanced, supporting the transition to more resilient, sustainable, and efficient energy systems. The future of BESSs lies in their ability to adapt dynamically to the complexities of energy demands and supply variations, paving the way for smarter grids and a more sustainable energy landscape.

To ensure the continued evolution and widespread adoption of BESS technologies, future research should focus on several key areas. First, enhancing safety and reliability is crucial, particularly for large-scale and second-life BESS applications. Research should prioritize the development of more advanced thermal management systems and fire suppression technologies, as well as better predictive models to understand battery degradation and performance over time. In addition, creating more durable and reliable storage solutions will be essential, especially as second-life batteries become more widely used. Second, addressing the lack of standardized regulations and interoperability is vital for scaling BESS deployment. Establishing global safety standards, performance benchmarks, and consistent regulatory frameworks will facilitate the smoother integration of BESSs into existing energy systems. Research on harmonizing international regulations for ESaaS and energy storage-sharing mechanisms will also be critical in fostering scalable, market-driven solutions.

Third, advancing AI and ML applications will be key to optimizing BESS operations. Further research should focus on the development of more sophisticated reinforcement learning algorithms tailored to BESS applications, enabling real-time adaptation to grid fluctuations and market conditions. These AI-driven approaches can improve energy arbitrage, demand response, and predictive maintenance, making BESSs more efficient and reliable. Additionally, integrating second-life batteries into BESS applications offers both environmental and economic benefits, but challenges such as performance variability and safe operation must be addressed. Research into BMSs will be essential for optimizing second-life battery performance and ensuring long-term reliability.

Finally, cost reduction remains a significant barrier to the widespread adoption of BESSs. Research should focus on lowering the capital costs of battery storage through innovations in battery chemistry, manufacturing techniques, and scalable production methods. Furthermore, the development of effective profit-sharing mechanisms for ESaaS and energy storage-sharing models will be crucial to ensure fair returns for all stakeholders and incentivize investment in BESSs. By addressing these key challenges and opportunities, future research can propel BESS technologies to the forefront of the global energy transition, enabling more resilient, sustainable, and efficient energy systems.