ASICS for Battery Management Systems
Lithium-ion battery packs are becoming increasing commonplace in a wide range of applications including automotive, marine, aerospace, rail and energy storage. However, whilst there have been many advances in lithium-ion cell chemistries in recent years, the design of the Battery Management System (BMS) that ensures the performance and safety of the pack has remained broadly the same.
A typical BMS consists of a wiring harness that measures the voltage of each parallel string of cells, i.e. cells connected positive-to-positive, negative-to-negative. Additionally, a number of temperature probes, such as thermocouples, will be present in the pack, but these will typically cover a large number of cells each. In a centralised BMS, all sense wires are directly connected to a master board, which checks for cells that are exceeding their maximum or minimum specified voltage or temperature and takes action to resolve the issue, for example by reducing the charge/discharge current or activating a cooling system. In a decentralised BMS, slave boards perform some calculations before sending data back to the master board. However, such systems are far from optimised for lithium-ion battery packs.
To explain the limitations of a conventional BMS, let’s analyse the BMS used on the world’s best-selling electric car. The Nissan LEAF has 192 cells in its battery pack, arranged 96 in series, 2 in parallel. With its conventional wired setup, the BMS master board only has access to 96 voltage readings, a maximum of 4 temperature readings and no cell-level current readings. If a cell in a parallel pairing starts to fail, the master board will have no indication of irregularities in current flow between the two cells, and any increase in temperature could take a considerable amount of time to reach the nearest temperature sensor. As a result, individual cells could undergo sub-optimal or abuse conditions and the vehicle would be completely unaware of the onset of cell failure until the cell is in a more advanced state of degradation.
A conventional BMS also has limitations when packs are being prepared for second life applications. When a module is disconnected from a battery pack, all data that is indicative of the State of Health of the module is lost. Furthermore, unless there are no parallel cells in the module, the BMS had no access to cell-level State of Health data to begin with. This results in labour- and time-intensive characterisation of modules in order to grade them for second life applications. The expensive and bulky wiring harness is most likely disposed of in the process. Some manufacturers have attempted to reduce this expense by daisy-chaining voltage sense data, but this risks communications from a large portion of the battery pack being lost should a single connection fail.
There is a clear need for the ability to measure cell-level data in battery pack, even on parallel cells; communicate that data to the master board in a simple, secure and inexpensive manner; and retain State of Health data with the cell throughout its lifespan. Based in Edinburgh, UK Dukosi a company which is leading the next generation of battery management has wiped the slate clean and built upon its sensor-on-a-chip expertise to deliver a completely new and fit-for-purpose solution to battery management. Their ASIC (Application Specific Integrated Circuit) can be embedded into a cell at the point of manufacture or attached to the top of an existing cell, where it will wirelessly relay raw cell data such as voltage and temperature, and calculated data such as State of Charge and State of Health, to the master board via a single proprietary near-field RF antenna. This eliminates 95% of the wiring harness and provides a breadth and depth of valuable data that was previously unavailable with a conventional BMS.
Returning to the Nissan LEAF example, a BMS based on the Dukosi cell monitoring system would give the master board access to all 192 voltage measurements, 192 temperature measurements, and 192 current, State of Charge and State of Health measurements. Additionally, the solution provides an on-cell event log that details the conditions that the cell has been subjected to during its working life. This gives the BMS full knowledge of the health and performance capabilities of every cell in the pack, improving the performance and lifespan of the pack and facilitating predictive rather than reactive maintenance, thus improving the availability of the vehicle. Each cell communicates its data back to the BMS master board via a single RF antenna, ensuring that every cell in the pack remains in contact with the master board even in the event of a single cell failure, unlike other communications protocols such as daisy chains. When a cell is removed from the pack, the cell-powered ASIC stays with the cell, as does all of the valuable data contained within it. This data can be accessed via an RFID reader within seconds, vastly decreasing the time taken to grade cells for second life applications and greatly increasing the throughput and safety of battery reuse and recycling facilities.
Another key advantage of the Dukosi system over a conventional BMS is that it is chemistry-agnostic and therefore intrinsically future-proof. Many vehicles have lifespans of several decades, but battery technology in 10, 20- or 30-years’ time will not be the same as battery technology today. Many upcoming cell chemistries have different cell voltages, meaning that the number of cells in series in a replacement battery pack for the vehicle would be different from its original pack, and potentially not a multiple of 8 or 12 as per the inputs on many conventional wired master and slave boards. A conventional BMS master board could not cope with this change, resulting in the need to completely retrofit the electronics within the vehicle, which is time-consuming and expensive. However, the Dukosi system is not limited to a specific number of voltage inputs and can handle any number of cells in series, so the master board would merely require a quick over-the-air software update to configure it for the new cutting-edge cell chemistry in the replacement battery pack. This will allow electric vehicles to continue to operate for decades, improving their range and performance with every replacement battery pack.
Excitingly, it transpires that this revolutionary new method of cell monitoring is also cost-effective. Extensive cost analysis performed by industry-leading experts, based on the BMS of a leading electric vehicle, indicate that the replacement of the wiring harness and slave boards with ASICs on every cell and a single wireless antenna provide a cost saving of 20% to the vehicle manufacturer. In return, the manufacturer gains access to a plethora of data never previously obtainable with a conventional BMS, and can improve the performance of their vehicle and lifespan of their battery pack.