In the rapidly evolving landscape of energy storage systems, lithium-ion batteries have become the predominant choice for large-scale energy storage battery applications due to their high energy density, long cycle life, and low self-discharge rate. However, the internal heat generation during charging and discharging processes poses significant safety risks, especially under high rate conditions or fault scenarios. Traditional monitoring methods, such as thermocouples and resistance strain gauges, suffer from electromagnetic interference, slow response, and limited multiplexing capability. To address these challenges, I have developed a novel approach based on fiber Bragg grating (FBG) sensors integrated into a pouch-type energy storage battery module. This work demonstrates the simultaneous measurement of temperature and strain at the individual cell level, enabling distributed and real-time condition monitoring for large-scale energy storage battery systems.
The core innovation lies in the packaging structure that decouples temperature and strain responses. By embedding two types of FBG sensors—temperature-only (T-FBG) and strain-sensitive (S-FBG)—between adjacent battery cells using an epoxy spacer, I achieved independent sensing of thermal and mechanical parameters. The T-FBG was mounted inside a ceramic tube using thermally conductive silicone rubber to isolate it from strain, while the S-FBG was pre-stretched and fixed in a groove directly on the battery surface. This configuration allows accurate measurement of both parameters without cross-sensitivity, which is critical for energy storage battery health assessment.
The principle of FBG sensing is based on the periodic modulation of the refractive index in the fiber core. When a broadband light source is launched into the fiber, the Bragg grating reflects a narrowband spectral component centered at the Bragg wavelength λB, which is given by the well-known equation:
$$ \lambda_B = 2 n_{\text{eff}} \Lambda $$
where neff is the effective refractive index and Λ is the grating period. Any external perturbation that alters neff or Λ—such as temperature change or mechanical strain—shifts the reflected wavelength. For a temperature-only effect, the wavelength shift is:
$$ \Delta \lambda = \lambda_B (\alpha + \xi) \Delta T = K_T \Delta T $$
where α is the thermal expansion coefficient, ξ is the thermo-optic coefficient, and KT is the temperature sensitivity. For strain, the shift is expressed as:
$$ \Delta \lambda = \lambda_B (1 – p_e) \Delta \varepsilon = K_\varepsilon \Delta \varepsilon $$
where pe is the effective photoelastic constant and Kε is the strain sensitivity. When both temperature and strain are present, the total wavelength shift is the sum:
$$ \Delta \lambda = K_\varepsilon \Delta \varepsilon + K_T \Delta T $$
This linear superposition allows decoupling when two FBGs with different sensitivities are employed, which is exactly the strategy used in my energy storage battery monitoring system.
To validate the sensor performance, I conducted calibration experiments. The temperature sensitivity of T-FBG was measured using a climate chamber (0–65 °C), yielding a linear response with R² > 0.99 and a sensitivity of 10.06 pm/°C. The strain sensitivity of S-FBG was determined by applying known loads and comparing with a reference strain gauge, giving a sensitivity of 1.7 pm/με. The temperature cross-sensitivity of S-FBG was also calibrated and found to be 26.84 pm/°C, which is used for compensation during simultaneous measurements. These results are summarized in Table 1.
| Sensor type | Measured parameter | Sensitivity | R² |
|---|---|---|---|
| T-FBG (temperature) | Temperature | 10.06 pm/°C | 0.998 |
| S-FBG (strain) | Strain | 1.7 pm/με | 0.995 |
| S-FBG (temperature cross) | Temperature | 26.84 pm/°C | 0.997 |
The experimental setup consisted of 16 pouch-type LiFePO₄ battery cells (25 Ah nominal capacity, 3.2 V cutoff) stacked in a module with epoxy spacers. Each cell was instrumented with both T-FBG and S-FBG sensors embedded in the spacer layer. The FBG array had wavelengths ranging from 1526 nm to 1563 nm with 1.5 nm spacing, allowing wavelength-division multiplexing. A commercial interrogator (ELT-ZD8, 1 pm resolution, 1 Hz sampling) recorded the spectra. The module was cycled using a battery tester (BT100V60AC12) under constant-current constant-voltage (CC-CV) charging at rates of 0.3 C, 0.4 C, 0.5 C, and 0.6 C, followed by 0.5 C constant-current discharge, with 3-hour rests between cycles. All tests were conducted at an ambient temperature of 20 °C inside a chamber to minimize external disturbances.
Figure 2 (not shown) presents the typical temperature and strain curves obtained from the FBG sensors during a complete charge-discharge cycle. The data from 14 functioning sensors (two were damaged during assembly) showed excellent consistency in trend, though minor numerical differences existed due to the position of each cell within the module. The charging phase (stage I) exhibited a gradual temperature rise, reaching a maximum at the end of CC-CV charging. The highest temperature occurred at the central cells (e.g., sensors T7 and T8), while edge cells (T1 and T14) showed lower values, with a temperature gradient of 3–4 °C between the center and the edge. After charging, during the rest period (stage II), temperature decreased, then rose again during discharge (stage III), and finally decayed during the final rest (stage IV). The maximum temperature rises during charging for 0.3 C, 0.4 C, 0.5 C, and 0.6 C were 6.3 °C, 7.9 °C, 9.2 °C, and 10.5 °C, respectively. During discharge at 0.5 C, the average temperature increase was about 6.0 °C across all cycles, confirming the repeatability of the energy storage battery module.
Strain data mirrored the temperature trends. During charging (stage I), the strain increased steadily, with average rates of 59, 90, 124, and 161 με/h for 0.3 C, 0.4 C, 0.5 C, and 0.6 C, respectively. The total strain change at the end of charging was 197, 225, 247, and 268 με, respectively. The strain increase is attributed to thermal expansion as well as volume changes due to lithium intercalation. During discharge (stage III), the strain decreased initially but then showed a slight increase near the end. The net strain change during discharge averaged 49, 55, 58, and 67 με for the four preceding charging rates (all discharged at 0.5 C). The consistent strain behavior across cycles further demonstrates the reliability of FBG sensors for energy storage battery condition monitoring.
Table 2 summarizes the average temperature and strain variations for different charging rates.
| Charging rate | Max temperature rise (°C) | Average temperature rate (°C/h) | Strain change (με) | Average strain rate (με/h) |
|---|---|---|---|---|
| 0.3 C | 6.3 | 2.2 | 197 | 59 |
| 0.4 C | 7.9 | 3.2 | 225 | 90 |
| 0.5 C | 9.2 | 4.3 | 247 | 124 |
| 0.6 C | 10.5 | 4.8 | 268 | 161 |
To verify the accuracy of the FBG sensors, I compared the measurements with a commercial NTC thermistor (accuracy ±0.2 °C) and a resistive strain gauge (DH3818Y, BE120-3AA) attached to a representative cell. The comparison, performed during a 0.6 C charge followed by 0.5 C discharge, showed almost identical profiles with minor deviations only at extreme points. This validation confirms that the proposed FBG-based system can reliably replace conventional sensors in energy storage battery modules, offering the added benefits of electromagnetic immunity and multiplexing.
The significance of this work lies in the ability to monitor temperature and strain at the single-cell level within a multi-cell energy storage battery module. The spatial distribution of thermal and mechanical loads can be used to identify early signs of abnormality, such as localized overheating or swelling. For instance, the observed 3–4 °C difference between center and edge cells indicates that thermal management systems must account for such gradients to prevent accelerated aging. Similarly, the strain evolution provides insights into the mechanical state of health, which is often overlooked in traditional battery management systems. By integrating FBG sensors directly into the module architecture, my approach paves the way for smarter and safer energy storage battery systems.
The experimental results also demonstrate a clear correlation between charging rate and both temperature and strain. Higher rates generate more heat and larger strain due to increased joule heating and faster lithium diffusion. This relationship can be modeled to predict the performance limits of the energy storage battery module. For example, the linear relationship between strain rate and C-rate (slope approximately 340 με/h per C-rate) can serve as a parameter for dynamic stress estimation.
In conclusion, this study presents a practical and robust method for real-time monitoring of temperature and strain in energy storage battery modules using FBG sensors. The novel packaging design effectively decouples the two parameters, and the sensor array provides distributed measurements across multiple cells. The system exhibits high sensitivity, excellent linearity, and good agreement with conventional sensors. The data acquired under various charging and discharging conditions offer valuable insights for battery thermal management, state estimation, and fault diagnosis. The FBG technology demonstrated here holds great promise for enhancing the safety and reliability of large-scale energy storage battery installations.