The Potential of Energy Storage and Advanced Battery Management for a Sustainable Future in Europe
Cost-effective energy storage is crucial to achieve the European Green Deal targets and improve the security of electricity supply in the EU, allowing greater flexibility in the grid and facilitating higher levels of renewable energy integration.
Revolutionising battery management systems (BMS) with advanced AI models and improved sensing technologies.
The cost-effective energy storage is a crucial element for achieving European Green Deal targets, clearly representing an enabler to contributing to the security of the electricity supply in the EU. Thus, it improves grid flexibility and allows higher penetration levels of renewable energy sources to create a decarbonised and more electrified society while contributing to the diffusion of distributed generation and following a sustainable and circular approach, for instance, by means of leveraging second-life batteries. In fact, a battery’s first life lasts between 10-15 years and is likely to retain more than two-thirds of its usable energy storage. Depending on their condition, used EV batteries can be repurposed for up to additional 10 years in “second-life cases” such as stationary energy storage, also known as the battery’s “second life”.
Furthermore, the requirements of the BMS are getting more and more advanced to maximise the performance of Liion batteries in usage. However, BMS obtains little information from a real battery, making it difficult to accurately indicate the ageing and safety status of a battery, and necessitates full life cycle management. Moreover, the on-board BMS cannot store or process large amounts of data during the operation of a vehicle, with poor real-time capability and data utilization rate. For efficient battery management, it is necessary to in-depth study the mechanisms, such as battery ageing and thermal runaway. Besides, the integration of advanced technologies like DT, artificial intelligence (AI) into the BMS is promising to realize battery life cycle data management. On the other hand, there are huge challenges in the research of accurate state estimation including smart charging, fast charging, thermal management, and extending useful life.
To develop and embed low-cost sensors which provide new physical information to the BMS
Creating and embedding new streams of physical data which can augment and enrich the information available to the BMS is crucial to enabling smarter BMS. The project will develop and evaluate low-cost temperature (CU) and ultrasonic (BATH) transducers and will take advantage of thermal cameras that can be embedded and are capable of feeding the BMS valuable information about cell internal and external temperature distributions, SoC and SoH. This information can be used by the BMS to better manage the battery, maximising its utilisation whilst mitigating any degradation.
To design a hardware abstraction layer platform
The Hardware Abstraction Layer (HAL) records and standardizes real and new sensor data from a battery prototype, ensuring error-free functionality. It’s scalable for various sensor arrays, allowing testing and comparison with simulations. It potentially serves as transducers to avoid BMS sensor interference and connects the physical demonstrator with the Digital Twin (DT).
To develop multiphysics modelling tools to continuously assess the SoX and RUL of Li-battery
Tracing internal battery structure changes aids understanding of aging, with consistent contributors even amid varied chemistries. Models are vital for State of Charge (SoX) and Remaining Useful Life (RUL) assessment. Battery changes stem from inputs like electricity and temperature, tied to distinct physical domains needing a multiphysics model. This approach, adaptable to diverse chemistries, operates within a multiphysics tool. It serves as a reference for data-driven BMS methods too.
To develop AI based models for explainable SoX prediction
Novel AI approaches for the simultaneous analysis of thermal images (simulated/real) and sensors will be introduced. The objective is to: (1) diagnose failure inside the battery pack using, an AI approach based on numerous sensors such as ultrasonic transducers and thermal cameras; (2) predict the battery SoX using a neural network and explain the degradation using post-hoc explainable models; (3) multimodal learning to combine image and time-series sensor data provided by transducers and temperature sensors to monitor battery health and pinpoint its degradation and failure reasons.
To design an innovative, connected and smart DT based BMS
The goal is to design an innovative and smart DT based BMS. A hybrid BMS, combining all the proposed explainable AI models, will be developed for that purpose. Through Edge computing capabilities, the connected BMS can compute physical data locally. In addition, Cloud computing could be considered for achieving computation with heavier models.
To make recommendations for future standard for predictive maintenance in the Cloud
ENERGETIC will identify all necessary information exported from batteries to assess concrete and abstract levels of performance and efficiency in order to prepare new services such as carsharing, fleet management or massive chains of energy storage systems to work in a predictive and optimised way. Information must be generalist, preparing the way to any kind of technology within energy storage systems, for any size and independent from brand.
To demonstrate and validate the ENERGETIC innovative smart DT based BMS
This will be made experimentally by accelerated ageing of cells at a specific SoH and after that to realise model assessment (ageing + safety) using different use case profiles (1 st life following automotive fleet use case) and 2 nd life (following stationary use case).
To facilitate the uptake and exploitation of ENERGETIC results by the academic community
technology developers and end-users, through targeted dissemination activities and link with the EU project and members of the Battery Partnership. This objective will pave the way for bringing the technology up to TRL9 on the long term.
Enjoy a new form of energy
Sequential Stages in Battery Tech, Modeling, and BMS Development
HOW ENERGETIC ANSWERS THESE MATTERS
Hence, the ENERGETIC project aims at developing the next generation BMS for optimizing batteries’ systems utilisation in the first (transport use case) and the second life (stationary use case) in a path towards more reliable, powerful, and safer operations. To do so, the ENERGETIC project contributes to the field of translational enhanced sensing technologies, exploiting multiple AI models, supported by Edge and Cloud computing. This will enable the path to future services based on data provided through the Cloud. ENERGETIC’s vision not only encompasses monitoring and prognosis of the remaining useful life of a Li-ion battery with a digital twin, but also encompasses diagnosis by scrutinising the reasons for degradation through investigating the explainable AI models.
