What happens when cells are not manufactured to the same quality standard or subject to cell grading and testing? Power management innovator Chris Hale of of Chimera Energy looks at the impact on battery packs of cell-to-cell variation and what the differences are between the grades.
It is well understood how lithium-ion batteries can be highly reactive and susceptible to thermal runaway events or even catastrophic failures. It is perhaps less well understood how significant cell-to-cell variations at the manufacturing level dictate overall battery life, reliability and safety. The impact of faulty batteries can not only inconvenience the user and pose potential health and safety risks but also end up costing and damaging a company’s reputation.
Understanding the risks and potential reliability issues that may be faced using lithium-ion cells, first comes from an understanding of the battery’s chemistry and failure modes, and secondly from understanding manufacturing process variations. The latter requires delving deeper into the mechanisms impacting reliability, accelerated degradation or potential premature failure and, for that, understanding potential issues during manufacture. For all the mechanisms influencing cell variation, a number of characteristic tests will be carried out, with thresholds dictating the eventual grade of the cell.
By better understanding the impact of potential defects, we can be more informed when selecting to opt for the safer A-grade cells, the riskier B-grade – or perhaps even the kamikaze route of choosing C-grade. Whichever route we take, we will still need to consider an appropriate level of testing to ensure a battery pack is balanced with cells of similar characteristics.
On the flipside, if you’re a consumer, how do you know the battery packs you’re purchasing will safely last the distance? Selecting a battery based on the manufacturer’s specification, warranty offering and battery price may be all that we have to go on, but how much can we really rely on this? Testing is generally costly if done thoroughly, but how much, or how little testing, is considered sufficient? Companies vying for competitiveness will be looking to keep costs as low as possible, but the question here would be ‘where are those cost savings made’? If it is in testing, it might beg the question of whether a cheap battery is the right and safe way to go, especially if you can’t guarantee corners haven’t been cut – or low-quality B-grade or even C-grade cells haven’t been used.
It may be fair to say that the intended application will likely dictate our bias on where we get our batteries. For example, reliability and safety are key factors if you’re buying a new electric car with a warranty good for 8-10 years or 100,000 miles. For an E-bike, however, it may simply be a question of cost, albeit with a warranty typically of just 12 months.
In retrospect, if you’re buying a battery good for 2,000 cycles, you might well expect better than a 5-year life. However, if the warranty period falls far short of the expected life of the battery, with all the disclaimers on usage, can we really trust that the cells will go the distance? In the event it does fail within the warranty period, be mindful that not all warranties are worth the paper they’re written on. BMS logging is a powerful tool to prove the one instance of ‘abuse’ rendering the warranty void by the time you come to claim.
To put this into perspective, I followed an interview with a warehouse operator using lead-acid batteries for materials handling applications. They reported that around 67% of their batteries failed before the end of the warranty period. Given the general day-to-day usage of the batteries, the warranty was almost always void for one reason or another! At the end of the day it is very easy to claim abuse, or use out of specification – more so in lithium-ion batteries with BMS logging facilities.
So where does this all fit in with testing? In this case, I’d say reputation. If you can buy from a reputable source likely to honour a warranty (with a warranty period that makes sense), you can be fairly sure they would have a reasonable process of cell quality testing. From the manufacturer’s perspective, understanding the causes of potential in-warranty failures helps to drive those failures down.
Opting to use B-grade cells because, for example, they offer a 30-40% cost saving on A-grade versions is fine if we understand why they are B-grade and why they are unlikely to offer the same life. Even for A-grade cells, not all will be manufactured to be 100% of the specification. There are plenty of factors providing sufficient cell-to-cell variation, promoting the need for adequate quality sorting. There may also be a different view on what qualifies as an A-grade or B-grade cell.
If we assume that not all manufactured cells are defect-free, the question would then be: what are the likely defects and to what degree do the defects impact on a cells performance, longevity and safety?
A good illustration of typical issues is depicted in Fig 1, showing the key issues during manufacture, identified as:
(A) The deflected copper current collector, is an example of cell component deformation.
(B) Burrs on the tab.
(C and D) Examples of impurity particles in the cathode and anode, respectively.
(E and F) High-resolution visualisations of the defective electrode regions with non-uniform cathode packing and delamination, respectively.
All of these defects could profoundly influence the battery’s performance in practical applications.
Regarding (A), the current collector (typically Cu), plays an important role in lithium nucleation/growth, local current density and lithium-ion flux distribution.
Burrs on the tab (B) will risk puncturing the separator, creating an internal short.
Regarding (C)&(D), depending on the composition, the metallic impurities can be directly involved in the chemical reactions. These particles could potentially alter the surface chemistry, leading to increased electrochemical impedance and polarisation.
If the impurities are located near a separator, then there will be a risk of penetrating the separator and creating an internal soft or hard short circuit.
Particle packing at the electrode level (E), plays a significant role in affecting the ageing and lifetime of the battery.
Delamination of electrodes (F) affects the electrochemical behaviour of the cell and is heavily dependent on the electrode adhesion and drying temperature.
Of course, each of these potential defect areas is generally introduced during the manufacturing process, and can arise from:
- Raw material processing: the purity of raw materials used within the anode and cathode will have a significant impact on cell performance (one of the current concerns for using recycled material)
- Mixing: both anode and cathode materials are mixed to give a level of uniformity, if the mixing isn’t thorough enough uniformity may be impaired
- Coating and drying: once mixed, the material is coated onto foil; however, uniformity cannot be guaranteed and the drying rates/time will also impact the crystalline structure
- Calendering: once coated and dried, the foil is passed between rollers to reduce the thickness; over-calendering can damage and distort the foil
- Slitting: as the foil tends to be up to 1m wide, precision cutting is required to the right size, the cut edges can lead to splits and notches or rough edges that risk piercing the separator leading to internal shorts and higher self-discharge
- Winding and filling: anode, cathode and separator material are wound and inserted into cans. At this point, the cells are quality tested before filling as the last step renders the cell live. Filling is a process of adding the electrolyte – injecting exactly the same volume. Ensuring even distribution between layers is not easy
- Formation: activation and cycling of the cells to identify voltage/capacity grading will identify initial B-grade cells
- Ageing: 28 days of ageing for stabilisation and testing of internal resistance, weight, size and appearance, grading further into A or B categories
If we assume that each manufactured cell will have a greater or lesser degree of potential defects, the question then remains of what impact does each aspect have on the life and operation of the cell? If we look at a typical battery over its life, we will generally see a fair spread of capacities between cells as they head towards the end of life (see Fig 2); the weakest cell being the limiting factor for the useable life of a battery.
During manufacturing, cells will be tested and graded, with all those passing acceptable thresholds being grade-A and those that don’t as grades B or C.
If we accept that there is a fair spread of relative capacity in aged A-grade cells, what can we expect from a spread of aged B-grade cells? Perhaps to understand this would be to look closer at the characteristics that can be impacted when defining what is A-grade and what is B-grade. Of course, when we buy a B-grade cell, we may not know what characteristic(s) it had failed.
A-grade vs B-grade
The A, B or C grading during the manufacturing of lithium-ion cells follows a process of evaluating cell characteristics that generally follow:
- Self-discharge rate: this is the highest risk as it represents the greatest potential for issues during use and storage – including a higher possibility of internal (soft/hard) short circuit
- Low capacity: easy to identify and, if grouped with other similar capacity cells, would be less of a risk to use
- Higher internal resistance: balancing will be an issue and the ability to supply high current (especially as the cell ages), and increased heat generation is also a potential issue
- Size/weight out of spec: may represent packing density issues with tight fit battery enclosures
- Voltage and capacity level mismatch: a cell’s open circuit voltage(OCV) represents its relative capacity, which should be consistent from cell to cell
- Stock storage duration: if an A-grade cell is on the shelf for too long before being sold, it may be downgraded to B-grade
C-grade cells, would typically have a much higher self-discharge rate or capacity fade and would (should) normally be destroyed or recycled.
The use of B-grade cells, which should have any QR code markings removed by the manufacturer to aid identification, can be fairly unpredictable.
A few OEMs and battery pack suppliers have experienced issues with using B-grade cells and cell packs because the batteries struggle to perform to datasheet expectations and fail within warranty periods. Fig 3 represents a comparison test of A & B-grade cells showing similar initial capacities, but with completely different capacity fade over cycling.
On the other hand, Fig 4a demonstrates a high self-discharge over a 400-hour period, which will impact greatly the capacity spread of cells in storage. The self-discharge rate will also be compounded at higher states of charge (SoC) when left in storage, especially if greater than 80% – as can be seen from Fig 4b.
Due to the variable characteristics that represent a substandard cell, identifying sub-standard cells on first inspection might prove challenging. For those buying cells off the web, beware of the ‘genuine’ <cell manufacturer> cells, they may be ‘genuine’ but not necessarily of a good grade. Aside from increased capacity fade (Fig 3), increased impedance and higher self-discharge rates (as demonstrated in Fig 4a) will also generate issues during storage. Both storage and operation of these batteries may well lead to an increase in ‘balancing’ issues for the BMS electronics, which could also impact on safety and rate of degradation for some of the cells, or simply cause the BMS to shut down with a ‘balancing’ fault.
So with all these potential issues, the method and degree of testing for each cell become more critical. The results from testing will of course generate a greater or lesser degree of useful information depending on the sophistication of the test for grading or qualifying cells for use in a battery pack.
To give an example, a simple test might involve just an OCV measurement as part of quality inspection. This test may determine a voltage spread amongst all the cells enabling grouping, based on voltage, to provide a more balanced battery pack. The spread of voltage may also indicate voltage/capacity mismatch or a greater degree of self-discharge. Self-discharge, of course, depends greatly on the length of time cells sit on a shelf before testing.
A simple voltage test will not tell you much about capacity variations, internal resistance or self-discharge rates.
AC/DC impedance tests
Also, an easy test to perform, either through an impulse load test or using more sophisticated electrochemical impedance spectroscopy (EIS) equipment. The internal resistance accounts for the voltage drop across the battery’s terminals when a load is connected, compared with no-load, and can be derived from OCV measurements.
Capacity tests
Determining the capacity of a cell is a bit trickier without doing a full charge-discharge cycle. Gaining a suitable indication of absolute capacity from quick and simple tests with any degree of accuracy is at the least, challenging.
Self-Discharge
The self-discharge rate of a cell is heavily dependent on SoC (Fig 4b) as well as temperature. Although there are several methods for determining self-discharge, given the very low rates of capacity loss over time, equipment accuracy and test duration become more significant. 10-14 days is often suggested, although there are those reporting reasonable indications of self-discharge rates from tests taking only 1-2 hours.
Voltage/capacity mismatch
OCV may well give an indication of SoC following a nominal curve, however for some cells the OCV at a given %SoC may deviate from the nominal profile. Testing requires knowledge of the capacity of a cell as described above, with challenges of accuracy for any shortened tests.
In summary, all lithium-ion cells will have a degree of variation, whatever grade they are. The grade of the cell and level of testing will ultimately impact the reliability, longevity or safety of a pack. Going cheap, where a manufacturer specifies long life but offers a short warranty, is something to consider carefully.
