Development of Cell Selection and Screening Method and Its Validation Process for the Multi-Combination Battery System

Abstract

A commercial-off-the-shelf (COTS) cylindrical 18650 Li-Ion cells have been proven to be extremely reliable power sources from many electronic applications for long periods and adopted to the EV applications. The cylindrical cells have been considered a reliable power source for the aerospace battery system, thanks to their high energy and power density. However, some electronic applications' field failure histories and some test programs have indicated that batteries built with cylindrical 18650 cells in multi-cell configurations (series and parallel) have experienced safety issues under various test conditions. Another requirement of the aerospace battery is a long-life cycle under harsh operating conditions. A tiny potential risky outlier cell, which cannot be identified and detected during the cell manufacturing process, can severely affect the multi-combination battery system's safety and reliability, like the EV, ESS, and Aerospace battery system. This paper studied a cell selection and screening method and developed a validation process to evaluate the efficiency and reliability of the cell selection and screening process. A cell selection method includes visual inspection, mass and dimension check, some electrical tests (OCV, CCV, AC Impedance), capacity measurement, and cell level safety test according to widely accepted cell safety standards and battery industrial best practices of the lithium-ion battery. The cell screening method and procedure followed the process and criteria on the NASA technical document (Specification for Acceptance Testing Commercial Lithium-Ion Cells). This program adopted the 6-sigma concept on the data analysis of the cell selection and screening process, which rejects the outlier cells of the 3-sigma level of each test item. The validation process was designed with two test conditions: 1) cycle test on the regular operation of the Human-spaceflight's mission operating conditions and 2) accelerate test condition of the maximum (most severe) cycle condition of the cell makers specification document. The regular operation cycle test was designed to evaluate the screening process's efficiency, whether the 3-sigma level screening is acceptable to the performance of the battery system using all passed cell samples from the screening. The regular cycle test selects four types of 4S string battery groups from different combinations of the cell samples. And it was performed according to the Human-spaceflight's mission operating conditions to validate the 4S can meet the required cycle condition. On the other hand, the accelerated cycle test was designed to identify the difference of cell degradation and capacity decay between the two types of cell combinations by the cycle test. One group use cells randomly selected samples from the non-screen cell, and the other group used selected cells from passed group cell as a control sample. Both 4S strings were cycled at the accelerate test condition, the maximum (most severe) cycle condition of the cell makers specification document. The battery capacity decay and the cell's internal resistance changes by DC pulse technics were checked regularly every 100 cycles during the cycle test.The performance of both 4S was almost similar until 200 cycles. However, the non-screen 4S string's capacity decay speed was much faster and reach death after the 500 cycles, while the screened 4S runs without any severe capacity decline until 800 cycles. EIS (Electrochemical Impedance Spectroscopy) technique was applied to each cell of the 4S string. The DPA (destructive Physical Analysis) and electron microscope analysis were used to identify the electrode's degradation degree and study the cell failure mechanism during the accelerated cycle tests. The DPA and electron microscope analysis (SEM-EDS) identified that gradual damage of the cathode electrode material of the non-screened 4S string during the accelerated cycle test was the leading cause of the cell capacity decay. This research discusses that non-screened 4S string may have different characteristic cell among the four cells in the 4 S string compare to the screened 4S string. That can initiate the stress to the cell and then increase the cell internal resistance, following enhanced cell heat generation and eventually getting the cell imbalance. The cell imbalance and heat generation get the cell getting worse after 200 cycles and eventually kill the cell and 4S sting. This research concludes that the used cell screening process is effective and beneficial as a cell screening process for the multi-cell battery system, especially for the long-term cycle-life battery system, like EV, ESS, and aerospace applications. Figure 1