Abstract
Lithium-ion batteries (LIBs) are pivotal for modern battery energy storage systems due to their high energy density, long lifespan, and cost-effectiveness. However, cyclic volume changes in electrodes during charge-discharge processes induce mechanical strain, which correlates closely with the battery’s state of charge (SOC). This study employs fiber Bragg grating (FBG) sensors to monitor strain evolution in graphite-lithium iron phosphate (AG||LFP) pouch cells. By designing a robust encapsulation method for FBG integration, we achieved in-situ strain measurement under operational conditions. Strain data revealed reversible mechanical behavior linked to phase transitions in electrodes, enabling SOC estimation with errors below 10% in plateau regions. These findings highlight FBG sensors as a promising tool for enhancing the safety and reliability of battery energy storage systems.
1. Introduction
Battery energy storage systems rely heavily on LIBs for applications ranging from electric vehicles to grid-scale storage. Despite their advantages, LIBs face challenges in mechanical stability due to electrode expansion/contraction during cycling. Such strain accumulation not only affects structural integrity but also provides critical insights into SOC and state of health (SOH). Traditional strain-monitoring methods, such as strain gauges or X-ray diffraction, suffer from electromagnetic interference, bulkiness, or impracticality in real-world settings. Fiber Bragg grating (FBG) sensors, with their compact size, multiplexing capability, and immunity to harsh environments, offer a transformative solution.
This work focuses on AG||LFP pouch cells, a mainstream choice for battery energy storage due to their thermal stability and affordability. We demonstrate how FBG sensors enable real-time strain monitoring, elucidate strain-SOC correlations, and propose a framework for SOC estimation superior to voltage-based methods.
2. Experimental Design
2.1 Battery Assembly
Commercial AG||LFP pouch cells (2 Ah capacity) were sourced from Hunan Cubic New Energy. Key specifications include:
| Parameter | LFP Cathode | Graphite Anode |
|---|---|---|
| Active material ratio | 96.7% LFP | 95.7% Graphite |
| Binder ratio | 1.8% PVDF | 2.9% PVDF |
| Electrode loading | 16 mg cm−2−2 | 7.8 mg cm−2−2 |
Electrodes were dried at 70°C for 14 hours and assembled in an argon-filled glovebox. Electrolyte (1M LiPF66 in EC:DEC (1:1) with 5% FEC additive) was injected, followed by vacuum sealing. Cells underwent three formation cycles (200 mA, 2.5–3.65 V) under pressure clamping.
2.2 FBG Sensor Integration
FBG sensors were calibrated for temperature and strain sensitivity using:ΔλB=KTΔT+KϵΔϵΔλB=KTΔT+KϵΔϵ
where and Kϵ=0.854 pm μϵ−1Kϵ=0.854pmμϵ−1. Two FBGs were deployed: one bonded to the cell surface, and another loosely placed for temperature compensation (Figure 1).
2.3 Electrochemical and Ultrasonic Testing
Cycling tests were conducted at 600 mA (2.5–3.65 V). Ultrasonic imaging (TOPS-LD50A scanner) assessed gas formation and structural integrity using 2 MHz transducers.
3. Results and Discussion
3.1 Strain Evolution During Cycling
FBG-measured strain exhibited three distinct phases during charging (Figure 2):
- Rapid Strain Rise (0–20% SOC): Lithium intercalation into graphite induces abrupt volume expansion.
- Plateau Phase (20–80% SOC): Strain stabilizes due to graphite’s phase transition (2L → 2 phase).
- Final Strain Surge (80–100% SOC): Complete lithiation drives further expansion.
Discharge showed reverse trends, with total strain reversibility (Δϵ≈387 μϵΔϵ≈387μϵ), confirming mechanical stability. Strain-temperature decoupling achieved <2.5 μϵμϵ error.
3.2 Electrode-Specific Strain Behavior
Embedding FBGs into single-layer cells revealed contrasting electrode dynamics:
| Electrode | Charging Strain Trend | Discharging Strain Trend |
|---|---|---|
| LFP Cathode | Linear decrease | Linear increase |
| Graphite Anode | Non-linear increase | Non-linear decrease |
LFP strain (Δϵ≈120 μϵΔϵ≈120μϵ) correlated with Li++ migration along crystallographic axes, while graphite’s larger strain (Δϵ≈250 μϵΔϵ≈250μϵ) dominated overall cell behavior.
3.3 SOC Estimation via Strain
Voltage-based SOC estimation falters in LFP’s flat voltage plateau (ΔV < 0.2 V). Strain, however, provided monotonic relationships during charging (Figure 3). A piecewise model segmented strain-SOC curves into four regions:SOC={a1ϵ+b1(0–20% SOC)a2ϵ+b2(20–80% SOC)a3ϵ+b3(80–100% SOC)SOC=⎩⎨⎧a1ϵ+b1a2ϵ+b2a3ϵ+b3(0–20% SOC)(20–80% SOC)(80–100% SOC)
Model validation showed <10% error in plateau regions, outperforming voltage-based methods (Table 1).
| SOC Range | Strain-Based Error | Voltage-Based Error |
|---|---|---|
| 0–20% | 12% | 25% |
| 20–80% | 8% | 35% |
| 80–100% | 15% | 20% |
4. Implications for Battery Energy Storage
- Enhanced Safety: Real-time strain monitoring detects mechanical anomalies (e.g., gas formation, electrode cracking) early, critical for grid-scale battery energy storage.
- Precision SOC/SOH Tracking: Strain signals complement voltage/current data, enabling robust battery management systems (BMS).
- Scalability: FBG multiplexing allows simultaneous monitoring of multiple cells, reducing deployment costs.
5. Conclusion
This study demonstrates FBG sensors as a transformative tool for battery energy storage systems. By correlating strain evolution with SOC, we address critical limitations of traditional voltage-based methods. Future work will integrate machine learning to universalize strain-SOC models, further advancing the safety and efficiency of LIBs in large-scale energy storage.