Semiconductors for Battery Systems 11.6 Percent More Usable Battery Energy with Chip-on-Cell Technology

Related Vendors

FAULHABER Drive Systems

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

SIMTOS (Seoul International Machine Tool Show)

The Scottish semiconductor company Dukosi derives significantly more usable energy directly from measurement data collected at the battery cell level. At the Pretzl Connect industry event, the company presented model calculations for electric vehicles and stationary storage systems.

To what extent does the measurement accuracy of a battery management system affect the amount of energy that can actually be utilized from a battery? Dukosi presented specific figures on this topic at the Pretzl Connect industry conference: In an example using LFP cells, the company’s Smart Cell architecture is said to increase usable energy by up to 11.6 percent, depending on the application.

To achieve this, Dukosi uses measurement ICs directly on the individual cells. The company contrasts the traditional multi-channel BMS with, among other things, short and uniform measurement leads, synchronized measurement values, and temperature sensing at each cell. In the calculation example shown, this significantly reduces the assumed error in state-of-charge estimation. The analysis is based on LFP cells, a Sigma-Point Kalman filter, and profiles for the UDDS urban driving cycle and grid stabilization via Fast Frequency Response (FFR).

Gallery

More Usable Energy instead of More Capacity

The 11.6 percent mentioned here does not mean that the cell physically stores more energy. Rather, Dukosi argues that more accurate measurement data reduces the uncertainty in determining the state of charge. This allows the battery management system to utilize the available energy with fewer reserves. In the UDDS example, Dukosi cites an effect of up to 11.6 percent; in the FFR profile of a stationary storage system, the figure is 6.2 percent.

Dukosi further calculates the potential economic impact for a battery storage system with 100 MW of power and 200 MWh of capacity. Based on the assumed 6.2 percent increase in usable energy, the company estimates an additional revenue potential of $7.8 million for the CAISO scenario used. A second calculation assumes that, for the same usable power, 12.4 MWh less capacity would need to be installed. Assuming a cost of $165 per kWh, this would correspond to approximately $2.1 million in reduced investment costs.

These are model calculations and not results from an actual 200-MWh storage facility in operation. Above all, the figures show the system-wide effect that Dukosi derives from smaller measurement and estimation errors.

DK-NFLNK Implements the Concept at the Module Level

At the same time, Dukosi is extending this approach to existing battery architectures. With DK-NFLNK, contactless near-field communication is expected to be usable even with conventional multi-channel AFEs. A system hub communicates with the host processor via SPI, while nodes are connected via SPI to the respective AFEs of the battery modules. Communication between the hub and the nodes takes place wirelessly via a so-called “black channel.”

This means a manufacturer does not have to switch immediately to Dukosi’s complete chip-on-cell architecture. The cells in a module remain connected to their AFE, while the DK-NFLNK replaces the communication link at the module level. Dukosi aims to reduce the number of cables and connectors while simultaneously creating a migration path to its own Smart-Cell architecture. A demonstrator combines a DK8502 node with an LTC6811 AFE in a twelve-cell module; for development purposes, Dukosi is showcasing a DK8503 evaluation board. 

Battery Development and Simulation

5 Approaches to Optimized Battery Designs