Thermal Runaway Suppression Capability of State-of-the-Art Coolant Fluids for Lithium-Ion Battery Applications

Abstract

With the increasing market uptake of hybrid/electric vehicles, along with growing electrification of transportation, battery safety is under worldwide scrutiny. Lithium-ion chemistries have become popular amongst battery manufacturers due to their superior performance and lifetime compared to lead-acid/nickel-metal-hydride based chemistries. However, due to the nature of the materials and their potential exothermic reactions, lithium-ion cells possess a high risk of thermal runaway related hazards. In this study we aim to investigate how aftermath of a thermal runway can be suppressed by employing coolant fluid. Experiments were conducted to investigate the viability of various coolant fluids as thermal runaway suppressants under external heating abuse conditions; on lithium-ion cells. The test apparatus (see attached image file – figure 1) was half filled with a coolant fluid and a 21700-type cylindrical cell at 100% state of charge (SOC), held vertically, such that half the cell was submerged. An external heating setup was employed via nichrome based special heating elements on the non-submerged half of the cell. The cell was heated with a known power value. Two types of cooling fluid were trialled using identical test set-ups, and 9 experiments were conducted in the following categories: 3 control (no fluid), 3 boiling fluid (having low boiling point in range of 70 to 80 °C) and 3 dielectric oil. There were 3 thermocouples attached to the surface of the cell (1 closer to the positive tab, 1 near the geometric centre, 1 closer to the negative tab). Another 3 K-type thermocouples with range till 1300 °C and accuracy of ±2 °C were employed to measure cell surrounding temperature, i.e. on the pressure cap present on the positive side of the cell, in the coolant fluid and another thermocouple in the air on the likely cell venting path. In the control experiments, the cells initially became internally short-circuited, then, their current interrupt devices (CIDs) were triggered and the cells eventually went into thermal runaway (see attached image file – figure 2). Temperatures in the order of 1100 °C were recorded. In the boiling fluid experiments, there was an order of magnitude increase in the duration to internal short-circuiting compared to no fluid (order of 1000 seconds compared to 100 seconds for control test cases) and the CID of every cell was partially triggered (see attached image file – figure 3). Some of the cells did not undergo thermal runaway. In those cases, temperatures up to 187.7 °C was observed. However, 1 cell went into full thermal runaway with the CID being fully triggered (temperatures up to 394.3 °C was observed). Throughout all the boiling fluid tests, the volume of fluid reduced continually. This might be detrimental to the usability of the fluid as a thermal runaway suppressant in case of longer duration experiments. In the dielectric oil experiments, the time to internal short-circuiting was similar to the boiling fluid and the CID was not triggered on any of the cells. No cells underwent thermal runaway and maximum cell surface temperatures were in the region of 160 °C. Throughout all the dielectric oil experiments, the volume of fluid remained constant. It was observed that both the boiling fluid and dielectric oil have strong potential as thermal runaway suppressants, with the dielectric oil completely preventing thermal runaway in any of the tested cells. This is despite the cell only being half submerged in fluid. A cooling/safety system centred on these types of fluids could manifest, perhaps as permanently submerged battery packs, for hybrid/EV automotive, aerospace and rail applications. One potential drawback for the deployment of fully submerged systems is the additional weight required. This could be mitigated in a number of ways, one being the integration of suppressant fluids into current cooling systems. It should be also noted that the time taken to internal short circuit was longer for the cases where boiling fluid was employed compared to where dielectric oil. However, in one of the cases, where boiling fluid was employed, the cell went into full thermal runaway. This could be due to the low boiling point and high volatility of the boiling fluid which makes it unreliable in an open system. In further work, cells connected in serial strings will be considered under further thermal safety scenarios. Along with Fourier-Transform Infrared Spectroscopy (FTIR) based gas analysis to investigate the effect of various failure modes such as nail penetration, overcharge, external heating on the gas generation at different SOCs and heating rates in submerged environment. Figure 1