Reliable Cell Behaviour Predictions for Real Life Conditions Get Started...
Read MoreVineet Dravid
Founder, CEO oorja
Reliable Cell Behaviour Predictions for Real Life Conditions
Introduction
As electrification moves into the fast lane, we face new technical and economic challenges.
Battery packs make up to 40% of the total EV cost, and Lithium-ion cells make up to 70% of the cost of the battery packs. The cells are, therefore, the most critical component of the electric vehicle. Yet they are the least understood!
This is because the behaviour of cells is governed by a complex set of physical processes, viz. electrochemical, thermal, and electrical. Although several experimental and theoretical studies have been conducted to understand these processes, the effect of their interplay is still not well understood. The behaviour of the battery is highly nonlinear, which means predictions based on only data are not reliable.
This leads to overengineering, which increases both costs and resource utilisation.
Since the safety of battery systems is paramount and battery life cannot be compromised due to the high cost of packs, companies today are investing heavily in lab testing of batteries, which is both expensive and time-consuming.
However, this accelerated testing under controlled conditions does not provide an accurate representation of battery behaviour in real life. Further, physics-based models need extensive parametrization and are computationally intensive, which adds to the development time.
oorja uses a hybrid approach, combining data and first principles physics-based modelling to deliver reliable and timely predictions of real-life battery behaviour.
The oorja Approach
The first step towards developing a model with high fidelity is estimating the correct set of material properties for the cell. oorja uses state-of-the-art incremental optimisation schemes to enable quick and accurate property estimation using HPPC data.
Once the properties are estimated, we use the same advanced optimisation techniques on as few as 300 cycles of data to predict the correct set of degradation parameters across different conditions for battery behaviour. Fig.2 shows the optimisation performed and demonstrates the accuracy of the predicted curve.
oorja has also developed tools to estimate the right drive and duty cycles and charging profiles specific to the vehicle specifications and usage conditions. Using these conditions as an input to the physical model constructed using data, oorja further employs its latest simulation approaches to predict battery behaviour in real-life conditions.
Fig. 4a depicts the drive cycle for a battery pack designed for a two-wheeler, delivering 3.3 kW of power. This drive cycle needs to be repeated for ~2.4 times to get the required DoD over which the cycling needs to be done. Fig.4b shows the predictions for life and range. As depicted in the figure, the range reduces significantly within 300 cycles, equivalent to the degradation over 6 months.
oorja has been proven to be up to 8x faster than the next fastest modelling tool for a similar accuracy. This also enables oorja users to simulate their batteries under diverse real-life conditions, providing more confidence in the delivered product.
To explore how you can use oorja to simulate accelerated testing, contact us at info@oorja.energy
Latest Blog Posts
From Inconsistency to Integrity: How Physics Informed Machine Learning Tackles EV Battery Challenges
From Inconsistency to Integrity: How Physics Informed Machine Learning Tackles...
Read MoreDeciphering Charging Dynamics: Part 1 Why Does It Take So Long To Charge My Battery?
Deciphering Charging Dynamics: Part 1 Why Does It Take So...
Read More