Back to search

MAROFF-2-Maritim virksomhet og offsh-2

Data-Driven Degradation monitoring and prediction of BATteries for Maritime ApplicatioNs

Alternative title: Data-Dreven Degraderingsovervåkning og prediksjon av BATterier for Maritime ApplikasjoNer

Awarded: NOK 4.8 mill.

Project Number:

311445

Project Period:

2020 - 2023

Organisation:

Location:

The project has investigated data-driven methods for estimating state of health (SOH) of maritime battery systems (lithium-ion) during operation, and hence to monitor battery capacity and degradation. Battery cycling tests have been performed in laboratory experiments, where several battery-cells have been continuously charged and discharged over an extended period of time, with regular check-ups and capacity measurements. This has given new insight in degradation of such batteries and provided data to be utilized by data-driven methods. The laboratory test data will be made publicly available for other researchers outside the consortium. Additionally, the project has collected operational data from several ship installations over several years. A number of different approaches to data-driven capacity estimation of batteries have been explored. These include models based on the complete operating history of the battery systems and models based on snapshots, as well as machine learning models that needs to be trained with relevant training data and models that do not need pre-training and hybrid methods partly based on physical knowledge. Several scientific publications have been published describing these investigations. Purely data-driven for reliable and accurate monitoring of state of health have turned out to be very challenging and there are benefits and challenges associated with the different methods, as outlined in project reports and publications. One of the main findings within the project is that pure data-driven methods might not be enough - unless enormous amounts of relevant training data are available - and that hybrid solutions combining data with physics-based models may be preferred. One such model is developed within the project, promoted as the preferred solution, that combines operational data with Coulomb counting and an equivalent circuit model (ECM) as well as look-up tables from comprehensive characterization tests. The project has demonstrated that independent verification of SOH can be achieved directly from operational data by utilizing such models, without the need for dedicated capacity tests where the ship needs to be taken out of service for about a full day once per year. This represents an important step forward that will make it easier for battery providers and ship owners to fulfill class requirements for ships relying on battery power for propulsion. A separate work package in the project has analysed the integration of a large battery system onboard an existing cruise ship and performed LCA as well as investigated other aspects of such hybrid solutions for cruise ships.

All ships encounter similar challenges when it comes to hybridizing their propulsion systems, particularly in the passenger shipping sector, within the context of the growing environmental debate, international legislation (IMO GHG Strategy), and evolving customer expectations. These challenges are relevant to both the operating fleet and the newbuilding programs of shipping companies. With the anticipated introduction of fuel cell systems, the importance of battery storage is expected to increase even further. Results from this project contribute to reduce the cost of battery energy storage systems for ships and hence promote uptake in the maritime industries. DNV has explored ways of utilizing sensor data from normal operations in ship classification of electric ships and for independent verification of state of health. This will contribute to make DNV a more attractive classification society for fully electric and hybrid ships in the future, and also contribute to make fully electric ships a more attractive alternative for ship owners, helping to promote zero-emission alternatives in shipping. DNV GRD has gained knowledge on maritime battery systems that will be exploited in future research projects, further promoting maritime green shipping solutions. Improved ways of collecting, sharing and analyzing maritime battery data has enabled Corvus Energy to develop their products and services, in particular with respect to the independent verification of battery capacity required by class societies. This also benefits their customers and other stakeholders. Improved condition monitoring of onboard battery systems, may also contribute to the safety of electric ships. Carnival Maritime is proactively exploring environmental and financial advancements for its managed ships and the broader fleet of Carnival Corporation. The ongoing commitment to sustainability, innovation and decarbonization has led to the evaluation of battery concepts across diverse ship classes. However, comprehensive details on financial considerations and the battery's State of Health throughout its lifetime are essential for well-informed decision-making. This applies to both retrofitting existing ships in the current fleet and incorporating batteries into newbuilding programs. The knowledge acquired from the project provides crucial information, contributing significantly to the operational efficiency of the fleet by enhancing both environmental and financial aspects. The projects outcomes are instrumental in optimizing the asset's lifetime, mitigating system failures, and maximizing the economic benefits of the technology. As a result, they not only justify, streamline, and reduce the costs associated with battery system applications but also facilitate their market introduction in the maritime sector. Data obtained from Fraunhofer ISE's laboratory tests will be made publicly available to support other researchers on battery degradation monitoring and related topics.

The use of battery systems for onboard energy storage is an attractive alternative for many shipping segments, both from an economic perspective with significant potential for cost savings, but also from an environmental, regulatory and societal perspective. However, the safety of battery-powered ships must be ensured. One critical aspect is the ability to deliver, at any time during operations, the power demand for safe and reliable propulsion, maneuvering and operation. Failure to do this may lead to intolerable accidents with severe consequences. Thus, a reliable estimation and prediction of the available energy stored in the battery at any time is of paramount importance. One of the objectives of this project is to develop data-driven methods for prognostics of battery systems and to provide means for verifying the battery state of health (SoH) based on real-time sensor measurements. Currently, such validation is based on an annual capacity test,which has several limitations. The test typically requires the ship to be taken temporarily out of service and it is believed that more accurate and reliable estimates of SoH can be obtained based on continuous sensor measurements so that variability in loads, temperatures and depth of discharge can be taken into account. Operational data are needed, and these will be collected in the project from ships in operation, and reliable and secure strategies for data collection, storage and sharing will be addressed. Moreover, additional insight will be obtained from laboratory testing under variable conditions. Focus will be on aspects related to battery systems for cruise ships, including battery lifetimes, replacement strategies, life cycle assessment and shore connection procedures. The project will set new standards for reliability and lifetime prognostics and deliver recommendations and give input to standards, recommended practices and class rules and main project results and findings will be reported.

Funding scheme:

MAROFF-2-Maritim virksomhet og offsh-2