State of Health

State of health (SoH) is a figure of merit describing the condition of a battery or energy storage system relative to its ideal, new state [8]. It is a critical metric in battery management systems (BMS) that quantifies the degree of degradation a battery has undergone, effectively indicating its remaining useful life and current performance capability compared to its original specifications [5]. SoH is distinct from, yet complementary to, the state of charge (SOC), which measures the immediate available energy; while SOC is analogous to a fuel gauge, SoH represents the long-term aging and wear of the battery's "engine" [5][6]. The accurate determination of SoH is fundamental to functional safety, performance prediction, and lifecycle management for any system reliant on rechargeable batteries, from consumer electronics to electric vehicles [7]. The core value of SoH lies in its expression as a percentage, where 100% represents a battery in its pristine, factory-fresh condition, and 0% typically indicates a battery that has reached its end-of-life criteria, such as being unable to hold a specified percentage of its nominal capacity [5][8]. Key parameters used to estimate SoH include capacity fade (the loss of total charge the battery can store), power fade (the reduction in maximum deliverable power), and increased internal resistance [1][4]. Estimation techniques range from direct measurement methods, like full discharge capacity tests, to advanced data-driven and model-based algorithms that infer health from operational data, though these can demand significant computational resources [4]. Mitigating factors that accelerate health degradation, such as excessive heat, are addressed through battery thermal management systems employing air, liquid, or phase change material-based cooling [2]. The significance of State of Health spans numerous applications. In consumer electronics, such as smartphones, it directly informs users about battery performance and the potential need for replacement [3]. In electric mobility and grid storage, accurate SoH estimation is essential for safety, reliability, warranty assessment, and determining residual value, particularly for second-life battery applications [7]. The development of robust SoH estimation methods is a key technology area for enhancing the real-time performance and longevity of battery management systems [3][4]. As such, SoH serves as a foundational parameter for the operational and economic optimization of modern energy storage systems across industries.

Overview

State of health (SoH) is a critical figure of merit that quantifies the present condition of a battery or energy storage system relative to its ideal, new state [14]. This metric serves as a fundamental indicator of performance degradation, remaining useful life, and overall system reliability, particularly in applications where failure carries significant safety or economic consequences. The SoH is typically expressed as a percentage, where 100% represents a battery in pristine, factory-fresh condition, and 0% indicates that the battery has reached its end-of-life criteria, such as the inability to deliver a specified percentage of its rated capacity or power [14]. Accurate SoH estimation is a cornerstone of modern battery management systems (BMS), enabling predictive maintenance, optimal system utilization, and ensuring functional safety in complex applications like electric vehicles (EVs) and grid storage.

Definition and Core Parameters

The SoH of a battery is not defined by a single universal parameter but is instead a composite assessment derived from several key measurable characteristics that degrade over time and use. The two primary and most widely used indicators are capacity fade and power fade (or increase in internal resistance).

Capacity Fade: This refers to the reduction in the total amount of charge a battery can store and deliver, measured in ampere-hours (Ah). End-of-life is often defined as the point where the battery's actual capacity falls to 70-80% of its nominal or beginning-of-life (BoL) capacity. For example, a lithium-ion battery with a BoL capacity of 100 Ah that currently delivers only 82 Ah has a capacity-based SoH of 82% [14].
Power Fade / Internal Resistance Increase: This reflects the battery's diminished ability to deliver high currents, directly linked to an increase in its internal resistance (measured in milliohms, mΩ). A higher internal resistance leads to greater voltage sag under load, reduced available power, and increased heat generation. Power fade is critical for applications requiring high bursts of energy, such as EV acceleration.
Other Contributing Factors: While capacity and resistance are primary, a comprehensive SoH assessment may also consider:
- Charge/discharge efficiency reduction
- Changes in open-circuit voltage (OCV) characteristics
- Self-discharge rate increase
- Thermal behavior and heat generation profile

Importance in Functional Safety and System Management

The accurate determination of SoH is intrinsically linked to functional safety, especially in automotive applications governed by standards like ISO 26262 [13]. This standard defines a risk-based approach to safety throughout the lifecycle of electrical and electronic systems in road vehicles [13]. A BMS that inaccurately estimates SoH can lead to dangerous failures, such as unexpected loss of propulsion or inability to meet power demands during critical maneuvers like highway merging. Therefore, robust SoH algorithms are a safety-related element whose development process must adhere to the rigorous requirements of ISO 26262, including specific Automotive Safety Integrity Levels (ASIL) [13]. Beyond safety, SoH is essential for:

Warranty and Remaining Useful Life (RUL) Prediction: Manufacturers use SoH to assess warranty claims and predict when a battery pack will require replacement.
Second-Life Applications: Determining if a retired EV battery (e.g., at 70-80% SoH) is suitable for less demanding secondary uses, such as stationary energy storage.
Optimal Charging Control: Advanced charging protocols can adapt based on SoH to minimize further degradation.

Degradation Mechanisms and Influencing Factors

Battery degradation is a complex electrochemical process influenced by operational patterns and environmental conditions. Key mechanisms include:

Solid Electrolyte Interphase (SEI) Layer Growth: A passive layer forms on the anode, consuming active lithium ions and increasing resistance. This is a primary cause of capacity fade in lithium-ion batteries.
Lithium Plating: At high charge rates or low temperatures, metallic lithium can deposit on the anode surface instead of intercalating, permanently reducing capacity and increasing risk of internal short circuits.
Electrode Active Material Loss: Mechanical stress from cycling can cause cracking and detachment of active material particles (like lithium cobalt oxide or graphite) from the current collectors.
Electrolyte Decomposition: Breakdown of the electrolyte solvent and salt can occur, especially at elevated temperatures or high voltages. The rate of these degradation processes is accelerated by several stress factors:
Temperature Extremes: High temperatures (>45°C) accelerate chemical side reactions, while low temperatures (<0°C) promote lithium plating and increase internal resistance.
Depth of Discharge (DoD): Regularly cycling a battery to a high DoD (e.g., 80-100%) causes more mechanical strain than shallow cycles, shortening cycle life.
Charge/Discharge Rate (C-rate): High currents generate more heat and increase overpotentials, accelerating degradation.
Time (Calendar Aging): Batteries degrade even when not in use, primarily through SEI growth, which is strongly temperature-dependent.

Estimation and Monitoring Techniques

SoH cannot be measured directly and must be estimated through various model-based, data-driven, or experimental techniques. These methods are broadly categorized:

Direct Measurement Methods: These involve periodic full characterization tests, such as a capacity test (full discharge at a low C-rate) or an internal resistance test (using AC impedance spectroscopy or DC pulse methods). While accurate, these are often impractical for online, in-situ estimation as they require taking the system offline.
Model-Based Estimation: These algorithms use electrochemical or equivalent circuit models (ECMs) of the battery. Parameters of these models (e.g., internal resistance, capacity in an ECM) are estimated in real-time from operational voltage, current, and temperature data using state observers like Kalman filters. The drift in these parameters from their initial values is then correlated to SoH.
Data-Driven/Machine Learning Approaches: These methods utilize large historical datasets of battery cycling to train algorithms (e.g., support vector machines, neural networks, Gaussian process regression) to map features extracted from operational data—such as voltage curves during constant-current charge, incremental capacity (dQ/dV) peaks, or temperature rise—to the battery's SoH. Other recent reviews in this area, such as that by Kabir et al., explore the expanding role of these advanced computational techniques.
Prognostic Methods for RUL: These extend SoH estimation to predict future degradation trajectories, often using particle filters or other filtering techniques to project when the battery will reach a failure threshold.

Role of Thermal Management in SoH Preservation

Battery thermal management is a critical protection method to maintain cell temperature within an optimal window, directly preserving SoH and ensuring safety [14]. Excessive temperature is a primary accelerator of all degradation mechanisms. Effective thermal management systems aim to keep battery pack temperatures typically between 15°C and 35°C during operation and charging. Common approaches include:

Air-Based Cooling: Forced air (passive or active) is a simple and cost-effective method, often used in early-generation EVs and consumer electronics. Its cooling capacity is limited, making it less suitable for high-performance applications.
Liquid-Based Cooling: This more advanced method uses a coolant (often a water-glycol mixture) circulated through cold plates or jackets in contact with the cells or modules. Liquid cooling offers superior heat transfer coefficients and more precise temperature control, making it the standard for most modern electric vehicles.
Phase Change Material (PCM)-Based Cooling: PCMs absorb large amounts of latent heat as they melt at a specific temperature, passively stabilizing battery temperature. They are often used in conjunction with other systems or in applications with intermittent high-power demands. By maintaining temperature below critical threshold levels, these systems directly mitigate the rate of SEI growth, electrolyte decomposition, and other thermally accelerated aging processes, thereby extending the battery's useful life and preserving its SoH [14]. The design of the thermal management system is itself a safety-related consideration, as its failure could lead to thermal runaway.

History

The concept of state of health (SoH) as a figure of merit for energy storage systems emerged from the broader field of reliability engineering and condition monitoring. Its historical development is intrinsically linked to the evolution of battery technology, particularly the rise of rechargeable systems requiring performance prediction and failure management.

Early Foundations and Pre-Lithium Era (Pre-1990s)

The fundamental need to assess the condition of a battery relative to its ideal, new state predates the term "state of health." Early methods for primary (non-rechargeable) cells were simplistic, often involving voltage checks under load. The advent of lead-acid batteries for automotive starting, lighting, and ignition (SLI) applications in the early 20th century introduced more systematic, though still indirect, condition assessments. Techniques focused on measuring specific gravity of the electrolyte with a hydrometer, which correlates with state of charge and, to a lesser extent, overall condition. For nickel-cadmium batteries, which gained prominence in the mid-20th century for portable electronics and aviation, capacity checks through full discharge cycles became a common, if time-consuming, method to gauge health. These early approaches established the core principle that a battery's usable life and performance could be quantified, but they lacked the sophisticated, in-situ estimation algorithms that would later define SoH. The formalization of failure analysis methodologies provided a critical theoretical backbone for later SoH concepts. Methodologies like Failure Mode and Effects Analysis (FMEA), developed systematically by the aerospace and military industries in the 1940s and 1950s, and later, more detailed approaches like Failure Mode, Mechanisms, and Effects Analysis (FMMEA), created structured frameworks for understanding how systems degrade and fail [15]. These methodologies allowed engineers to derive the specific failure modes and mechanisms for electrochemical systems, moving beyond simple observation to predictive modeling of degradation pathways.

The Lithium-Ion Revolution and SoH Formalization (1990s-2000s)

The commercialization of lithium-ion battery technology by Sony in 1991 marked a pivotal turning point. These batteries offered significantly higher energy density but introduced new, complex degradation mechanisms sensitive to factors like temperature, charge voltage, and current. Managing these expensive and sensitive power sources in consumer electronics, and later in nascent electric vehicles, necessitated a more precise and continuous metric of condition. The term "state of health" began to be used consistently in technical literature during this period to describe this metric, typically expressed as a percentage [14]. It was commonly defined as the ratio of a battery's current maximum capacity or power capability to its nominal, beginning-of-life value [14]. This era saw the development of the first Battery Management Systems (BMS), which integrated voltage, current, and temperature monitoring. Early SoH estimation within BMS units was often rudimentary, relying on simple cycle counting or lookup tables based on operating conditions. However, academic and industrial research intensified, exploring model-based and data-driven approaches. The recognition that degradation was not linear and was influenced by complex electrochemical processes drove the search for more accurate estimators. The foundational understanding that excessive temperature is a primary accelerator of all degradation mechanisms led to parallel advancements in battery thermal management. Engineers developed protection methods to maintain temperature below critical thresholds, employing various cooling strategies including:

Air-based cooling (forced convection)
Liquid cooling (cold plates, coolant jackets)
Phase change material-based cooling

Algorithmic Advancements and Standardization (2010s)

The 2010s witnessed an explosion in research focused on SoH estimation algorithms, driven by the global push for electric vehicles and large-scale stationary energy storage. The limitations of simple cycle counting became apparent, prompting a shift towards methods that could adapt to real-world usage patterns. Key research directions included:

Equivalent Circuit Model (ECM)-based filtering techniques (e.g., Kalman filters) that used voltage and current data to track parameter changes like internal resistance. - Incremental Capacity Analysis (ICA) and Differential Voltage Analysis (DVA), which involved analyzing the shape of charge/discharge curves to identify degradation modes. - Machine learning approaches that trained models on large datasets to correlate operational data with capacity fade. A significant methodological advancement was demonstrated in research on battery modules, which showed that differential voltage analysis could be effectively applied when paired with a local data symmetry method to handle cell-to-cell variations [3]. Experimental validation of models also became more rigorous. For instance, one study combining a mixture Weibull reliability model with an equivalent circuit model reported an average Root Mean Square Error (RMSE) of 8% for capacity estimation, highlighting both the progress and the ongoing challenge of achieving high accuracy [15]. Concurrently, the industry began to standardize definitions and test procedures. The establishment of end-of-life criteria, often linked to a specific SoH threshold, was crucial for warranty, valuation, and repurposing decisions. The practice of defining end-of-life at 70-80% of beginning-of-life capacity became common for many applications, creating a clear link between the SoH metric and practical battery lifecycle management.

Integration and Real-World Application (2020s-Present)

In the current decade, SoH estimation has moved from a research topic to a critical, embedded feature in billions of devices. Its calculation is now deeply integrated with other BMS functions like state-of-charge estimation and thermal management. A prominent public example is Apple's performance management system for iPhones, introduced in 2017, which explicitly uses a combination of device temperature, battery state of charge, and battery impedance (a key indicator of power fade) to dynamically manage performance and prevent unexpected shutdowns in aged batteries. This implementation brought the concept of battery SoH and its real-world implications to mainstream consumer awareness. The focus has expanded from mere estimation to prognostic health management (PHM), aiming to predict remaining useful life (RUL). Furthermore, accurate SoH assessment is now the cornerstone of the second-life battery market, where batteries retired from electric vehicles, having reached an end-of-life threshold for automotive use, are graded and redeployed for less demanding applications like stationary energy storage. Recent comprehensive reviews of the field, such as that by Kabir et al., synthesize the vast array of estimation techniques—from electrochemical models to deep learning—and highlight the ongoing challenges of achieving robust, computationally efficient, and universally applicable SoH algorithms across diverse battery chemistries, formats, and usage profiles. The history of SoH continues to be written as it evolves from a diagnostic metric into a central element of intelligent energy storage system optimization and circular economy practices.

It is a critical metric for determining a battery's remaining useful life, safety, and suitability for its intended application. Unlike the state of charge (SoC), which indicates the present available energy relative to the current maximum capacity, SoH reflects the long-term, irreversible degradation of the battery's fundamental performance characteristics [5]. Accurate SoH estimation is therefore essential for predictive maintenance, warranty validation, and ensuring the reliability of systems ranging from portable electronics to electric vehicles and grid-scale energy storage.

Quantifying Degradation and Failure Modes

While capacity fade and power fade are primary indicators, SoH is fundamentally linked to the underlying physical and chemical degradation mechanisms within a cell. These mechanisms are complex and interdependent; some usage patterns and operating conditions lead to rapid degradation by one or more processes, and the interplay between mechanisms is still not well understood [2]. To systematically address these risks, the failure modes and mechanisms for any system can be derived using different methodologies like failure mode effects analysis (FMEA) and failure mode methods effects analysis (FMMEA) [2]. These analytical frameworks help identify potential points of failure, such as:

Internal short circuits due to separator breakdown or lithium dendrite growth
Loss of active lithium inventory through side reactions at the electrodes
Structural degradation of electrode materials, leading to increased internal resistance
Electrolyte decomposition and gas generation

Understanding these specific failure modes is crucial for designing accurate SoH estimation algorithms and implementing effective battery management strategies [13].

Measurement and Estimation Techniques

Direct measurement of SoH often requires controlled laboratory conditions. Techniques like Galvanostatic Cycling with Potential Limitation (GCPL), Chronopotentiometry (CP), and Constant Current (CC) cycling are suitable for precisely measuring a battery's actual capacity or internal resistance by performing a full charge or discharge cycle under controlled parameters [5]. However, these methods are intrusive, time-consuming, and impractical for online estimation during normal operation. Consequently, significant research focuses on indirect, data-driven estimation methods that infer SoH from readily measurable operational data. Advanced techniques include the use of Gaussian process regression, which combines mean and covariance functions to provide probabilistic estimates of SoH along with uncertainty bounds [4]. Other model-based and data-driven approaches analyze features such as:

Incremental capacity (IC) or differential voltage (dV/dQ) curves, where shifts in peak positions and amplitudes correlate with degradation
The shape and duration of constant-current charging voltage profiles
Electrochemical impedance spectroscopy (EIS) measurements at various frequencies

For instance, one method estimates the state of health of battery modules via differential voltage analysis with a local data symmetry method, leveraging the characteristic voltage plateaus during phase transitions in electrode materials [14]. These approaches allow for non-intrusive, real-time monitoring without taking the battery offline.

Integration with Battery Management and System Protection

A Battery Management System (BMS) uses SoH information to optimize performance and ensure safety. Building on the foundational role of BMS discussed earlier, functional safety standards like ISO 26262 for automotive applications mandate rigorous processes for hazard analysis and risk assessment, directly linking SoH monitoring to system safety goals [13]. The BMS adjusts operational limits based on SoH; for example, a degraded battery with high internal resistance may have its maximum charge and discharge current derated to prevent excessive heating and voltage sag. Battery thermal management is also a critical protection method to maintain the temperature below threshold levels, and includes air, liquid, and phase change material-based cooling systems [2]. As noted earlier, excessive temperature accelerates degradation, so thermal management works in concert with SoH-aware algorithms to prolong life. Furthermore, the sizing of electrical wiring and circuit breakers in a battery system must be based on the apparent (VA) power, which is inherently influenced by the battery's internal impedance—a key component of its state of health [6]. A practical example of SoH-informed performance management is found in consumer electronics. iPhone performance management works by looking at a combination of the device temperature, battery state of charge, and battery impedance [3]. As the battery ages and its impedance rises (indicating a lower SoH), the system may dynamically manage peak power demands to prevent unexpected shutdowns, directly linking measured battery health to user-experienced device performance.

Challenges and Future Directions

Accurate SoH estimation remains challenging due to the variability in battery manufacturing, the complexity of coupled degradation mechanisms, and the diverse real-world usage profiles a battery may experience [2]. Current research, including other recent reviews in this area such as Kabir et al., explores hybrid models that combine physical electrochemical principles with machine learning algorithms to improve adaptability and accuracy across different battery chemistries and operating conditions. The ongoing development of these techniques supports broader goals of battery longevity, second-life applications, and sustainable energy storage, reflecting a continued commitment to sharing knowledge about lithium battery technology and the critical functions of battery management systems [13].

Significance

State of Health (SoH) serves as a critical figure of merit for evaluating the condition of a battery or energy storage system relative to its ideal, new state [20]. Its significance extends far beyond a simple diagnostic metric, influencing economic decisions, safety protocols, system reliability, and the feasibility of sustainable second-life applications. While often expressed as a percentage, the interpretation and acceptable thresholds for SoH are not universal but are deeply contingent on battery chemistry and the specific demands of the application [16].

Beyond a Universal Metric: Application-Dependent Criticality

The conventional definition of SoH as the ratio of current available capacity to initial capacity provides a standardized baseline [19]. However, the criteria for determining a battery's end-of-life (EoL)—often linked to a specific SoH threshold—is an area of significant nuance. As noted earlier, a common EoL benchmark is when capacity fades to 70-80% of its beginning-of-life value. This benchmark, while useful for general comparisons, is an overly conservative definition when applied universally across all use cases [16]. Different applications possess varying tolerances for performance degradation. For instance, a battery pack in a stationary energy storage system for grid support may continue to provide valuable service at a SoH well below 70%, as its operational demands for power density and specific energy are less stringent than those of an electric vehicle. In contrast, an automotive application might require retirement at a higher SoH to maintain vehicle range, acceleration, and safety margins. Therefore, the true significance of a given SoH value cannot be assessed without considering the operational context, highlighting the need for application-specific health assessment frameworks [16].

Enabling Accurate State Estimation and Predictive Maintenance

Accurate SoH knowledge is fundamental to the reliable operation of complex battery systems, particularly in Battery Electric Vehicles (BEVs). It is intrinsically linked to the accurate estimation of the State of Charge (SOC), which indicates the remaining usable energy. SOC cannot be measured directly and must be inferred from measurable parameters like current, temperature, and voltage [18]. The accuracy of SOC estimation algorithms is highly dependent on having a precise and up-to-date model of the battery's health, as degradation alters key internal parameters like capacity and internal resistance. Building on the concept of capacity fade discussed previously, data-driven algorithms, including those employing statistical and machine learning techniques, have become prominent for predicting both SOC and SoH in BEV applications [20]. These methods leverage the strong data processing and nonlinear fitting capabilities of modern algorithms to model complex degradation behaviors from operational data. Advanced estimation techniques, such as the adaptive unscented Kalman filter, have been developed to simultaneously estimate SOC and SoH with verified accuracy under diverse operating conditions, which is crucial for reliable range prediction and battery management [17].

Informing Second-Life Economics and Circular Economy

The application-dependent nature of SoH is the cornerstone of the battery second-life industry. A battery pack retired from an automotive application at, for example, 75% SoH may no longer meet the rigorous demands for vehicle range and power but retains substantial value for less demanding secondary uses. As mentioned earlier, such packs are suitable for stationary energy storage applications. The economic viability of repurposing these batteries hinges entirely on a trustworthy assessment of their residual SoH and a prediction of their remaining useful life in the new context. Accurate SoH evaluation prevents the premature recycling of valuable assets and enables the creation of a cost-effective, circular economy for energy storage, reducing environmental impact and resource consumption. Recent reviews in this area, including work by Kabir et al., underscore the importance of robust SoH estimation methodologies in developing viable second-life pathways [16].

Challenges in Degradation Modeling and Future Outlook

Despite its importance, accurately modeling and predicting SoH remains a formidable scientific and engineering challenge. The parameterization of electrochemical and empirical degradation models requires extensive, long-term testing data that accounts for complex, interacting stress factors like cycling depth, rate, and environmental conditions [21]. The international battery community recognizes that improving these models is a major priority for advancing battery technology [21]. Furthermore, while data-driven methods show great promise, their performance is contingent on the quality and breadth of the training data. Physics-informed neural networks represent a cutting-edge approach that seeks to combine the flexibility of data-driven models with the foundational principles of electrochemistry, aiming for more stable and generalizable degradation modeling and prognosis [19].

Integral Role in Safety and Thermal Management

SoH is a vital parameter for system safety. Degradation often leads to increased internal resistance, which in turn generates more heat during operation for the same power output. Therefore, monitoring SoH trends provides an early warning for potential thermal risks. Battery thermal management systems (BTMS)—including air, liquid, and phase change material-based cooling—are critical protection methods designed to maintain cell temperature below critical thresholds [22]. The design and control strategy of a BTMS must account for the changing heat generation characteristics of a battery as its SoH declines. A comprehensive understanding of SoH allows for adaptive thermal management that ensures safety throughout the battery's entire lifecycle, from first use to potential second-life applications.

Applications and Uses

State of Health (SoH) is a critical metric that bridges fundamental battery science with real-world engineering and economic decisions. While typically expressed as a percentage, SoH is commonly defined as the ratio of a battery's current maximum charge capacity to its nominal beginning-of-life capacity [18]. However, this capacity-based criterion, while universal, is not inherently based on the specific application requirements of the system in which the battery operates [18]. The practical application of SoH spans from ensuring the safety and performance of primary-use systems like electric vehicles to enabling the burgeoning second-life market and informing sustainable end-of-life strategies.

SoH in Primary Applications: Electric Vehicles and Portable Electronics

In primary applications, particularly Battery Electric Vehicles (BEVs), accurate SoH knowledge is paramount for functional safety, performance optimization, and user experience. A precise understanding of SoH allows the vehicle's Battery Management System (BMS) to enforce operational limits that prevent hazardous conditions like overcharge or over-discharge, which become more likely as a battery degrades [14]. Furthermore, SoH is directly linked to the vehicle's driving range. As noted earlier, capacity fade reduces the total energy available. Without accurate SoH estimation, the BMS cannot reliably calculate the remaining usable energy, leading to either a limited effective range (if the system is overly conservative) or the risk of unexpected stranding (if the estimate is too optimistic) [18]. This estimation challenge is compounded by the fact that a battery's available capacity can also vary with temperature and discharge rate. To address this, modern BMSs employ sophisticated, data-driven algorithms to predict not only the State of Charge (SOC) but also the SoH in real-time. These algorithms often use adaptive filters, such as the Adaptive Unscented Kalman Filter (AUKF), which can simultaneously estimate SOC and SoH by processing measurements of voltage, current, and temperature while accounting for model uncertainties and noise [17]. The accuracy of these algorithms is crucial for preventing battery oversizing—a common design compromise where a larger, heavier, and more expensive battery is installed to guarantee a minimum range throughout the vehicle's life, which negatively impacts efficiency and cost [18]. For portable electronics, such as smartphones and laptops, SoH estimation informs the device's power management system, providing users with more accurate battery life predictions and managing charge termination to minimize further degradation. In all these primary uses, the continuous monitoring of SoH provides essential data for warranty validation, residual value assessment, and predicting the onset of end-of-life for the primary application.

Enabling the Second-Life Battery Market

As lithium-ion batteries proliferate in electric vehicles and renewable energy systems, managing their post-vehicle life cycle has become a significant challenge and opportunity [16][19]. When a battery pack reaches its end-of-life for automotive use—a threshold often defined, as mentioned previously, at 70-80% of its original capacity—it may retain substantial value for less demanding secondary applications [16]. Accurate SoH assessment is the cornerstone of this second-life market. The process involves:

Detailed SoH characterization at the module or cell level to identify units suitable for repurposing [16]
Sorting and regrouping cells with similar SoH and impedance profiles to ensure balance and safety in the new pack [16]
Re-certifying the reconfigured pack for its new duty cycle, such as stationary energy storage for renewable energy integration or backup power [16]

However, the decision-making process for determining which specific degradation pathways and SoH levels are suitable for which second-life applications is not yet well understood [16]. Clear technical standards, safety protocols, and economic models for second-life implementation are still largely lacking, creating a barrier to large-scale commercialization [16]. Developing universally accepted SoH grading standards is a critical research and policy need to unlock this market's potential, which is vital for improving the overall sustainability and economics of battery technology [16][19].

SoH for Prognostics, Health Management, and End-of-Life Planning

Beyond instantaneous assessment, SoH trends are used for prognostics and health management (PHM). By analyzing the trajectory of SoH decline over time and under specific usage conditions, algorithms can predict the remaining useful life (RUL) of a battery. This predictive capability is increasingly powered by data-driven and physics-informed models. For instance, Physics-Informed Neural Networks (PINNs) integrate fundamental physical laws of degradation with operational data to create more stable and reliable long-term prognosis models, even with limited early-cycle data [19]. These prognostic models enable:

Condition-based maintenance, where servicing is performed based on actual degradation rather than fixed schedules
Fleet management optimization for electric vehicle or grid storage operators
Informed end-of-life planning, allowing for the timely decommissioning and scheduling of battery collection for recycling or repurposing

End-of-life strategies are a direct application of SoH understanding. A comprehensive SoH history helps determine whether a battery is best suited for direct recycling to recover valuable materials like lithium, cobalt, and nickel, or if it has sufficient life remaining to be economically and safely deployed in a second-life application [16][21]. This decision is critical for developing a circular economy for batteries, mitigating environmental impact, and reducing reliance on virgin materials [16][19].

Integration with State of Charge Estimation and System Design

SoH is not an independent parameter; it is deeply intertwined with the accurate estimation of State of Charge (SOC). As a battery ages, its capacity decreases and its internal resistance increases, which alters the relationship between voltage, current, and stored energy. Therefore, accurate SOC estimation is dependent on an accurate, real-time SoH value. Advanced joint estimation algorithms, such as Dual Extended Kalman Filters (DEKF) or the aforementioned AUKF, are designed to solve this coupled problem simultaneously [17][18]. These filters continuously update both SOC and SoH estimates, using the coulomb count (current integration) and voltage response models that are parameterized by the latest SoH value [17][18]. This integration fundamentally influences system design. Knowledge of typical SoH degradation rates allows engineers to design the BMS, thermal management system, and electrical architecture with appropriate margins and adaptive strategies. It informs decisions about cell selection, pack sizing, and the specification of power electronics to ensure safety and performance are maintained throughout the battery's intended service life, accounting for its anticipated health decline [18][14].