Thesis
Model-free non-invasive health assessment for battery energy storage assets
- Creator
- Rights statement
- Awarding institution
- University of Strathclyde
- Date of award
- 2023
- Thesis identifier
- T16695
- Person Identifier (Local)
- 201689783
- Qualification Level
- Qualification Name
- Department, School or Faculty
- Abstract
- Increasing penetration of renewable energy generation in the modern power network introduces uncertainty about the energy available to maintain a balance between generation and demand due to its time-fluctuating output that is strongly dependent on the weather. With the development of energy storage technology, there is the potential for this technology to become a key element to help overcome this intermittency in a generation. However, the increasing penetration of battery energy storage within the power network introduces an additional challenge to asset owners on how to monitor and manage battery health. The accurate estimation of the health of this device is crucial in determining its reliability, power-delivering capability and ability to contribute to the operation of the whole power system. Generally, doing this requires invasive measurements or computationally expensive physics-based models, which do not scale up cost-effectively to a fleet of assets. As storage aggregation becomes more commonplace, there is a need for a health metric that will be able to predict battery health based only on the limited information available, eliminating the necessity of installation of extensive telemetry in the system. This work develops a solution to battery health prognostics by providing an alternative, a non-invasive approach to the estimation of battery health that estimates the extent to which a battery asset has been maloperated based only on the battery-operating regime imposed on the device. The model introduced in this work is based on the Hidden Markov Model, which stochastically models the battery limitations imposed by its chemistry as a combination of present and previous sequential charging actions, and articulates the preferred operating regime as a measure of health consequence. The resulting methodology is demonstrated on distribution network level electrical demand and generation data, accurately predicting maloperation under a number of battery technology scenarios. The effectiveness of the proposed battery maloperation model as a proxy for actual battery degradation for lithium-ion technology was also tested against lab tested battery degradation data, showing that the proposed health measure in terms of maloperation level reflected that measured in terms of capacity fade. The developed model can support condition monitoring and remaining useful life estimates, but in the wider context could also be used as the policy function in an automated scheduler to utilise assets while optimising their health.
- Advisor / supervisor
- Stephen, Bruce
- Resource Type
- DOI
- Date Created
- 2023
Relazioni
Articoli
| Thumbnail | Titolo | Data caricata | Visibilità | Azioni |
|---|---|---|---|---|
|
|
PDF of thesis T16695 | 2023-10-04 | Pubblico | Scaricare |