Electronic Theses and Dissertations Archive

Date

2026

Document Type

Thesis

Degree Name

Master of Science

Department

Mechanical Engineering

Committee Chair

Alexander Headley

Committee Member

David Rosewater

Committee Member

Jeffrey Marchetta

Committee Member

Yong Lee

Abstract

Battery energy storage systems are increasingly central to modern electric grids as they provide peak shaving, loadshifting, and frequency regulation. Lithium-ion batteries dominate the grid-scale scene due to their high energy density, long cycle life, and cost effectiveness. However, these cells degrade with use, losing capacity over time. When battery degradation is not accounted for in technoeconomic planning, simulations overestimate long-term energy availability and underestimate cost, with these errors compounding over the life of the project. Despite this, many planning tools do not include a degradation model. This work develops a lightweight, data-driven degradation model for integration into Sandia National Laboratories’ QuESt 2.0 technoeconomic planning software, built entirely from publicly available cycling data from BatteryArchive.org. The primary contribution to this work is a stress factor analysis applied to three lithium-ion chemistries, lithium iron phosphate, nickel manganese cobalt oxide, and nickel cobalt aluminum oxide, all from BatteryArchive.org. An exponential capacity decay model adapted from the semi-empirical framework of Xu et al. is used to estimate a per-amp-hour degradation rate for each cell. This rate is normalized to a selected reference condition to produce a dimensionless stress factor that quantifies how much faster or slower a given operating condition degrades a cell relative to the reference. Temperature, discharge rate, and depth of discharge are each evaluated independently across all three chemistries. For lithium iron phosphate, higher temperatures and discharge rates were identified as the dominant stress factors, each doubling the observed per amp-hour degradation rate compared to the reference. For all three chemistries, using a narrower depth of discharge window consistently reduced degradation per unit of throughput. Percentile based prediction intervals were also added, each derived from the 5th and 95th percentile of recorded degradation rates to help capture cell to cell variability. These give planners a best- and worst-case capacity projection, along with a median, for their selected chemistry.

Comments

Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to ProQuest/Clarivate.

Notes

Open Access

Download

Share

COinS