The pursuit of global carbon peaking and carbon neutrality goals has placed electrochemical energy storage, particularly lithium-ion battery technology, at the forefront of the energy revolution. As a critical carrier for electrical energy, the battery energy storage system plays an indispensable role in balancing the intermittency and volatility of renewable energy sources like wind and solar power, thereby significantly enhancing the stability, economy, and safety of power grids. Among various electrochemical storage technologies, lithium-ion batteries dominate the landscape of new energy storage installations. Within this domain, Lithium Iron Phosphate (LFP) batteries have garnered widespread adoption in large-scale battery energy storage system applications due to their intrinsic safety, long cycle life, and cost-effectiveness. However, the performance and longevity of a battery energy storage system are not solely determined by the electrochemical properties of individual cells but are profoundly influenced by their integration into modules and packs.
A key mechanical phenomenon affecting this integration is the swelling force generated during the charge and discharge cycles. This force originates from the reversible and irreversible volume changes of active materials within the cell as lithium ions shuttle between the cathode and anode. In a constrained environment, such as within a tightly packed module, these volume changes translate into significant mechanical stress. This swelling force is a critical parameter for evaluating the structural integrity, long-term reliability, and safety of a battery energy storage system. Excessive or unevenly distributed swelling forces can lead to accelerated degradation, connector failure, and in extreme cases, compromise the safety of the entire system. Therefore, a comprehensive understanding of the swelling force characteristics of LFP batteries, particularly their evolution over the full state-of-charge (SOC) range and throughout the battery’s lifetime (state-of-health, SOH), is paramount for the robust mechanical design of modules and packs. This study focuses on investigating these characteristics for a high-capacity LFP cell and its integration into different module configurations, providing essential data and insights for the safety-centric design of LFP-based battery energy storage system.
Experimental Methodology
The core of this investigation revolves around a commercially available, high-capacity prismatic LFP battery cell. The key specifications of this cell, which forms the basic building block of the studied battery energy storage system modules, are summarized in the table below.
| Parameter | Value | Unit |
|---|---|---|
| Chemistry | LiFePO4 / Graphite | – |
| Nominal Capacity | 280 | Ah |
| Nominal Voltage | 3.2 | V |
| Dimensions (T x W x H) | 72.0 x 173.7 x 207.2 | mm |
| Standard Charge/Discharge Power | 448 (0.5P) | W |
To simulate real-world application scenarios within a battery energy storage system, these individual cells were assembled into two distinct module configurations: a 1P12S module (12 cells in series, referred to as Module A) and a 1P8S module (8 cells in series, referred to as Module B). These modules were then placed into a specialized steel fixture designed to apply a constraint and measure the resulting swelling force. This fixture, equipped with calibrated force sensors, mimics the mechanical boundary conditions a module would experience when tightly packed inside a battery energy storage system enclosure.
The testing protocol was designed to evaluate swelling force behavior under both static (SOC variation) and dynamic (cycling aging) conditions. Initially, a preload force of approximately 300 kgf was applied to the module within the fixture to simulate the initial clamping condition in a pack. All tests were conducted in a controlled temperature environment of (25 ± 2) °C. The modules underwent continuous charge-discharge cycling at a constant power of 0.5P (3584 W for Module B and 5376 W for Module A). Each cycle consisted of a constant-power charge to an upper voltage limit (3.65 V per cell), a 30-minute rest period, a constant-power discharge to a lower voltage limit (2.5 V per cell), and another 30-minute rest. The swelling force was continuously monitored and recorded by the sensors throughout this process, allowing for correlation with real-time SOC and cycle number.
Results and Discussion
2.1 Swelling Force Variation Within a Single Cycle
The swelling force exhibited a distinct and repeatable nonlinear trajectory during a complete charge-discharge cycle. The results from an initial cycle are emblematic of this behavior. During the charging phase, the force increased almost linearly from the preload value to a first local peak at approximately 30% SOC. Following this peak, the force slightly decreased, reaching a trough around 60-70% SOC. It then resumed a strong linear increase, culminating in the absolute maximum force for the charge phase at 100% SOC. During the subsequent rest period, a slight relaxation in force was observed. Upon discharging, the force initially decreased from the rested 100% SOC value. It reached a minimum around mid-SOC before increasing again to form a third distinct peak near 30% SOC during discharge. Finally, the force decreased to its minimum value for the cycle as the cell approached 0% SOC. Crucially, the end-of-discharge force was higher than the initial preload force, indicating an irreversible thickness increase after just one cycle. The characteristic peaks are labeled for subsequent discussion: C1 (charging ~30% SOC), C2 (charging 100% SOC), and D1 (discharging ~30% SOC).
| Phase | SOC Region | Swelling Force Trend | Key Peak |
|---|---|---|---|
| Charge | 0% → ~30% | Linear Increase | C1 |
| Charge | ~30% → ~70% | Gradual Decrease | Trough |
| Charge | ~70% → 100% | Linear Increase | C2 (Max during charge) |
| Rest | 100% SOC | Relaxation (Decrease) | – |
| Discharge | 100% → ~60% | Decrease | – |
| Discharge | ~60% → ~30% | Increase | D1 |
| Discharge | ~30% → 0% | Decrease | Cycle Minimum |
2.2 Mechanistic Origins of Swelling Force Peaks
The observed force profile is a direct mechanical manifestation of the phase transformations occurring within the electrode materials during lithium (de)intercalation. The overall cell reaction can be represented as:
$$ \text{LiFePO}_4 + 6\text{C} \ \xrightleftharpoons[\text{Discharge}]{\text{Charge}} \ \text{Li}_{1-x}\text{FePO}_4 + \text{Li}_x\text{C}_6 $$
The graphite anode undergoes staged intercalation, where lithium atoms insert between graphene layers in specific sequences defined by the staging number \( n \), which indicates the number of graphene layers between adjacent lithium layers. The evolution proceeds through several distinct stages (dilute, Stage IV, Stage III, Stage IIL, Stage II, Stage I), each with a characteristic unit cell expansion along the c-axis. The staging phenomena lead to the characteristic voltage plateaus and, more importantly for this study, nonlinear volume expansion. The most significant expansion steps occur during the transition from Stage II to Stage I (near full lithiation) and during the early stage transitions (e.g., from Stage IV to Stage III/II). The LFP cathode, in contrast, undergoes a continuous, nearly linear contraction in unit cell volume upon delithiation.
The coupling of these two behaviors explains the tri-modal force peaks:
- Peak C1 (~30% SOC Charge): Corresponds to significant graphite expansion during early-stage transitions (e.g., IV→III→II), while the LFP cathode has only contracted a small amount. The net effect is substantial cell swelling.
- Trough (~60-70% SOC Charge): In this range, the graphite may be in a mixed or constant-slope staging region (like Stage IIL) with minimal volume change, while the LFP cathode continues to contract linearly. The cathode contraction temporarily counteracts anode expansion, leading to a net reduction in swelling force.
- Peak C2 (100% SOC Charge): Dominated by the large final expansion of graphite during the Stage II to Stage I transition, where the c-axis lattice parameter increases significantly, leading to maximum cell thickness and force.
- Peak D1 (~30% SOC Discharge): During discharge, the process is reversed. At around 30% SOC, the graphite is transitioning from a high-stage to a mid-stage compound (e.g., beginning of de-lithiation from Stage II), causing expansion, while the LFP cathode is only partially re-lithiated and thus still relatively contracted. This asymmetry during the hysteresis of the phase transformations results in the D1 peak.
The force relaxation during rest periods is attributed to the viscoelastic creep of internal cell components (like the separator and electrodes) and the fixture system under constant strain.
2.3 Evolution of Swelling Forces with Cycle Life (SOH)
The long-term cycling test revealed a critical evolution in the swelling force characteristics. While all three force peaks (C1, C2, D1) increased in magnitude with cycling due to cumulative irreversible expansion (from SEI growth, gas generation, etc.), their relative magnitudes changed significantly as the battery’s State of Health (SOH) degraded.
In early life (high SOH), the maximum force in a cycle was consistently the C2 peak (100% SOC). However, after approximately 500 equivalent cycles (corresponding to an SOH of about 94%), the C1 peak (charge 30% SOC) began to rival and then surpass the C2 peak. Beyond roughly 800 cycles (SOH ~92%), the hierarchy permanently changed: C1 > D1 > C2. The C2 peak (100% SOC) eventually became the smallest of the three peaks in a cycle during the later stages of life.
This shift can be understood by considering the loss of active lithium (Li+ inventory). As SOH decreases, the maximum amount of lithium that can be cycled is reduced. Therefore, at a degraded SOH of 90%, the “100% SOC” state in terms of usable capacity does not correspond to the same absolute lithiation level of the graphite as it did at Beginning of Life (BOL). The graphite anode at end-of-charge in a degraded cell is less fully lithiated—it may only reach a state equivalent to, for example, 90% SOC of a fresh cell. Referring to the volume coupling model, the expansion at this corresponding lithiation level is lower than that at the true 100% SOC (Stage I) of a fresh cell. Meanwhile, the expansion associated with the early-stage transitions (which cause peak C1) remains relatively prominent and becomes the dominant source of stress. This mechanistic understanding is vital for predicting the worst-case mechanical load in a battery energy storage system throughout its operational lifespan.
| Cycle Life Phase | Approx. SOH | Dominant Swelling Force Peak | Hierarchy (Force Magnitude) |
|---|---|---|---|
| Early Life (BOL) | 100% – ~94% | C2 (100% SOC Charge) | C2 > C1 > D1 |
| Mid Life | ~94% – ~92% | Transition from C2 to C1 | C1 ≈ C2 > D1 |
| Late Life | < ~92% | C1 (~30% SOC Charge) | C1 > D1 > C2 |
A key finding for system design is the relationship between the maximum cycle swelling force and SOH after this transition. For SOH values below approximately 90%, the maximum force (now primarily from peak C1) exhibits a strong linear correlation with decreasing SOH. This linear relationship can be expressed as:
$$ F_{max} = k \cdot \text{SOH} + b $$
where \( F_{max} \) is the maximum swelling force, SOH is expressed as a decimal (e.g., 0.9 for 90%), and \( k \) and \( b \) are constants. For the tested modules, linear regression yielded:
$$ F_{max, 1P12S} = -7429 \cdot \text{SOH} + 7532.9 \quad (R^2 = 0.997) $$
$$ F_{max, 1P8S} = -7402 \cdot \text{SOH} + 7305.8 \quad (R^2 = 0.996) $$
The near-identical slopes indicate that the rate of force increase with degradation is intrinsic to the cell chemistry and design, and is not significantly altered by the number of cells in series within the module. However, the total force magnitude is additive; the 1P12S module consistently experienced higher absolute forces than the 1P8S module under the same SOH condition. For instance, at 70% SOH, the predicted maximum force for the 1P12S module was approximately 2365 kgf. This additive effect is a critical consideration when scaling up from cell to module to pack in a battery energy storage system.
2.4 Simulation Analysis for Module Structural Design
Utilizing the experimentally derived force-SOH relationship, a finite element analysis (FEA) was conducted to assess the structural integrity of the module components under end-of-life swelling forces. Detailed 3D models of the modules, including end plates, steel binding straps, busbars, cells, and insulating pads, were created. The material properties (Young’s modulus, Poisson’s ratio, yield strength) were assigned accordingly.
| Component | Material | Yield Strength (MPa) |
|---|---|---|
| End Plate | A380 Aluminum | 159.0 |
| Busbar | AL-1060-O Aluminum | 27.6 |
| Steel Strap | SUS201 | 1168.0 |
| Bolt | 40Cr Steel | 525.0 |
The simulated load was the predicted swelling force at 60% SOH (e.g., ~2775 kgf for the 1P12S module, minus the initial preload). The boundary conditions fixed the module base to simulate mounting within a battery energy storage system rack.
The simulation results provided valuable insights:
- Displacement: The maximum displacement occurred at the center of the end plates, with values of 1.42 mm and 1.57 mm for the 1P12S and 1P8S modules, respectively. This deformation must be accounted for in pack design to prevent interference.
- Stress on End Plates & Busbars: The maximum stress in the aluminum end plates was found at reinforcing rib locations but remained safely below the material’s yield strength (131.7 MPa < 159 MPa). The busbars experienced stress concentrations at their ends due to the module’s bulging deformation, with values slightly exceeding the yield strength of the soft aluminum alloy (up to 28.0 MPa vs. 27.6 MPa). This highlights the importance of using designed busbars (e.g., with arched shapes) that can accommodate several millimeters of displacement without risking fracture, thus ensuring electrical connection reliability.
- Stress on Steel Straps and Bolts: The high-strength steel straps and bolts showed stress levels well within their safety limits, confirming their suitability for restraining the long-term swelling forces in a battery energy storage system module.
This simulation validates that the module’s mechanical design, when informed by empirically derived swelling force data, can withstand the anticipated loads over its target lifecycle, ensuring the structural safety of the battery energy storage system.
Conclusion
This study provides a comprehensive analysis of the swelling force characteristics in large-format LFP battery modules, offering critical insights for the mechanical design of battery energy storage system. The swelling force is not static but dynamically evolves with both the instantaneous SOC and the long-term SOH of the battery.
The force within a single cycle exhibits three characteristic peaks (C1, C2, D1) resulting from the coupled phase transformations in graphite and LFP materials. Importantly, the hierarchy of these peaks changes over the battery’s lifetime due to active lithium loss. After SOH degrades below approximately 92%, the peak during charging at around 30% SOC (C1) becomes the dominant source of mechanical stress, rather than the peak at full charge. Furthermore, beyond 90% SOH, this maximum swelling force shows a strong linear correlation with decreasing SOH, a valuable relationship for predictive maintenance and design life validation. The total force in a module scales with the number of cells in series, but the rate of force increase with degradation is an intrinsic cell property.
Finite element simulation, based on these experimental findings, demonstrates that with appropriate design—such as reinforced end plates, high-strength restraints, and compliant busbars—the module components can safely endure the swelling forces expected over the full operational life of a battery energy storage system. This work underscores the necessity of considering long-term mechanical evolution in addition to electrochemical performance when designing reliable and safe energy storage solutions. The methodologies and correlations established here serve as a foundation for optimizing module and pack designs, ultimately contributing to the enhanced durability and safety of grid-scale lithium-ion battery energy storage system.