New technologies often lead to “innovative blind spots,” and this is happening with electrification in batteries. Lithium-ion is a winning chemistry to store electrical energy but there is fear of failure that often starts in a faulty cell that propagates and engulfs the battery in fire. Such incidents are known with e-bike batteries and other devices.
Cadex is developing a technology that reads the electrochemical evidence of a battery with a frequency scan and displays the results in a Nyquist plot. Figure 1 demonstrates Nyquist plots taken from 100 high-quality 8650 Li-ion cells with Spectro Explore.
The Spectro Explorer services Li-ion and lead acid batteries from 3V to 48V with capacities of up to 300Ah. The test is minimally-invasive with a frequency scan from 2,000Hz down to 0.1Hz that takes about 30 seconds; longer for large batteries. Applications for the Explorer are quality control in manufacturing, examining uniformity of incoming batteries, and battery validation before shipment. Best results are achieved with preparations such as a prior full charge, followed by a short rest.
Dormant lead acid batteries develop sulfation, a deficiency that can be removed with exercise if the sulfation is soft. Hard sulfation that occurs during long dormancy at low charge can be permanent.
Primary lithium batteries also develop passivation while in storage as illustrated in BU-701: How to Prime Batteries (Figure 2). The passivation layer is removable with a brief discharge. These are added services that a Nyquist scan can discover.
What is a Nyquist Plot?A Nyquist plot consists of a resistive reading (Real Z) that is positioned on the horizontal axis, and a reactive analysis placed on the vertical axis. The resulting signature reflecting battery characteristics is divided into migration at high frequency on the left of the scale; charge transfer in mid-range; and diffusion at low frequencies on the right. Figure 2 illustrates a Nyquist plot serving as “golden sample” with adjustable pass/fail envelopes to set acceptance levels for quality control in manufacturing and performance check before deployment. |
Figure 2: Nyquist serves as golden sample with acceptance tolerances |
Figure 3 compares Nyquist plots with different battery performances, Figure 4 overlays a Nyquist plot of a good battery with one of low capacity and Figure 5 illustrates a battery with dendrite growth. Figure 6 compares Nyquist signature of aged lead acid batteries with sulfated packs. Unless otherwise noted all readings were taken at the Cadex labs in Canada and Germany.
|
|
|
|
Lead acid batteries stored for six months or longer develop sulfation; a dormancy effect that reduces performance but is reversible by exercise if done in time. Figure 7 shows the Nyquist plots of a lead acid battery that had a low charge transfer with a capacity reading of 75% when first serviced. After exercise with a charge/discharge cycle, charge acceptance improved and the capacity increased to 92%. Other lead acid batteries tested with various dormancy effects had similar improvements. More studies are needed in how the Nyquist can identify sulfation that lead to restoration |
|
Testing E-Bike BatteriesE-bike batteries are failing and in some cases catching fire. However, handled correctly, Li-ion is safe but the chemistry is less forgiving than lead acid and nickel-based chemistries. BU-304a and BU-304b advise how to keep Li-ion safe. |
|
We now scan batteries with the Explorer in an entire pack, as well as in parallel pairs. Figure 8 shows an open e-bike battery; Figures 9, 10 and 11 show the test results. With acquired knowledge, a Nyquist plot can easily be deciphered by technical staff. The task is simplified by comparing a faulty pack with a good one, knows as golden sample. As Figure 9 illustrates, a normal pack has a small footprint that is contained in the user-adjustable Acceptance Field, while a faulty battery steps outside set limits. The shapes of the “cat tails” may one day lead to identifying the nature of defect. |
|
Figure 10 shows an entire pack with one cell shorted. The graph also does a comparison with a parallel pair in which all cells are good as shown in the small oval. Adding one bad apple changes the order. Protection circuit and series connections add non-reactive resistance that shift the Nyquist plot of the pack to the right. |
|
Mixing non-reactive resistance with the reactive component of a battery is further exemplified in Figure 11. Here the protection circuit and long cables move the Nyquist plot to the right while retaining the signature of the Nyquist plot. The shift reflects the pure resistive part of the battery that will also assist in troubleshooting. |
Blue: Protection circuit bypassed Red: Includes protection circuit through connector with short cable Green: Includes protection circuit with long cable |
Conclusion
Testing batteries by EIS is not new, but scientists predict that future battery diagnostics rests in EIS analysis. Evaluation by Nyquist using the Spectro Explorer has the potential of becoming a household name, led by versatility, ease of use and low cost. Using these technologies will reduce innovative blind spots to maintain safety as the world electrifies with batteries. Typical EIS applications include:
According to the UL Research Institute, thermal runaway is a primary cause of fire in Li-ion. |
|