Lithium-Ion Battery Degradation in EV Duty Cycles

ISEF Category: Energy: Sustainable Materials and Design

Subcategory: Energy Storage  ·  Difficulty: Intermediate  ·  Setup: Home Setup  ·  Time: 1 to 2 Months

The Hook

Your phone battery does not just age because of time. It ages because of how you use it. The same idea matters for electric vehicles, where stop-and-go driving can stress a battery more than smooth highway travel. You can model that stress with real data and see which driving pattern is harsher.

What Is It?

This project studies how lithium-ion batteries wear down under different driving patterns. A duty cycle is just a usage pattern, like city driving, highway driving, or mixed stop-and-go travel. In a battery model, you can treat each pattern like a workout plan and ask which one causes more strain.

PyBaMM is a free Python tool that simulates battery behavior. You feed it a battery model and a usage pattern, and it estimates things like voltage, state of charge, and degradation trends. Think of it like a weather forecast for a battery, except the forecast changes based on how you drive.

NASA Ames battery datasets give you real cycling data to compare against your simulation. That makes your project more than a toy model. You can test whether your simulation matches measured battery aging and then explore how irregular EV-style use changes the results.

Why This Is a Good Topic

This is a strong science fair topic because you can test a real engineering question with public data and open-source software. Battery wear affects EV range, cost, and safety, so your work connects to a real-world problem. You can learn modeling, data cleaning, graphing, and validation without needing a physical battery lab. That makes the project realistic for a student, but still deep enough for serious research.

Research Questions

• How does irregular stop-and-go duty cycling change predicted lithium-ion battery degradation compared with steady driving patterns?
• What is the effect of different EV-style duty cycles on modeled voltage fade over time?
• Does a more aggressive acceleration and braking pattern increase state-of-health loss faster than a smoother pattern?
• To what extent does the choice of battery model in PyBaMM change degradation predictions for the same duty cycle?
• Which summary features of a drive cycle, such as high current peaks or frequent rest periods, best explain the simulated aging trend?
• How does model prediction compare with the NASA Ames battery dataset across different cycling conditions?

Basic Materials

• Laptop or desktop computer with Python installed.
• Internet access to download the NASA Ames battery dataset and PyBaMM documentation.
• Python package manager such as pip or conda.
• PyBaMM library.
• Jupyter Notebook or another Python notebook environment.
• Spreadsheet software such as Google Sheets or Excel.
• Digital notebook for tracking model settings, assumptions, and outputs.

Advanced Materials

• Laptop or desktop computer with Python installed.
• Access to the NASA Ames battery dataset in raw file format.
• PyBaMM with optional solver and parameter set extensions.
• Jupyter Notebook or VS Code with Python support.
• NumPy, pandas, matplotlib, and SciPy.
• ImageJ or similar plotting tool for figure cleanup if needed.
• External hard drive or cloud storage for versioned outputs.
• Git or another version control tool.

Software & Tools

• PyBaMM: Simulates lithium-ion battery behavior and degradation under custom usage patterns.
• Python: Runs the model, processes dataset files, and automates comparisons.
• Jupyter Notebook: Lets you document code, graphs, and notes in one place.
• pandas: Cleans and organizes battery time-series data.
• matplotlib: Creates clear plots for voltage, current, and degradation trends.

Experiment Steps

1. Define the battery question you want to answer, then choose one drive pattern variable to change first.
2. Select a PyBaMM degradation model that matches your skill level and the kind of output you want to compare.
3. Map the NASA Ames dataset into the same format as your simulation inputs so the two can be compared fairly.
4. Build a baseline run with one clean duty cycle, then add irregularity to test how the model responds.
5. Plan controls that keep battery type, parameter set, and comparison metrics fixed while you vary only the duty cycle shape.
6. Decide how you will judge agreement between simulation and data, such as error, trend match, or rank order of aging.

Common Pitfalls

• Mixing different NASA battery cells in one comparison, which makes the degradation trend hard to interpret.
• Changing several drive-cycle features at once, which hides which pattern actually caused the effect.
• Comparing simulation output to raw data without matching units, time steps, or state-of-charge definitions.
• Treating every battery fade curve as linear, which can distort model fit and make the error look smaller than it is.
• Ignoring parameter sensitivity, which can make a model look accurate for the wrong reason.

What Makes This Competitive

A class-level version of this project just plots battery fade. A stronger version tests more than one duty cycle, compares multiple degradation metrics, and checks whether the same conclusion holds across battery models. You can also add sensitivity analysis to show which driving features matter most. That kind of careful comparison turns a simulation project into real engineering research.

Project Variations

• Compare city, highway, and mixed-duty cycles to see which one produces the fastest predicted battery aging.
• Test how changing the frequency of rest periods affects degradation in the same base drive cycle.
• Use a different public battery dataset and see whether the same duty-cycle pattern produces the same aging ranking.

Learn More

• NASA Battery Data Set: Search the NASA Prognostics Center of Excellence pages for the Ames battery datasets and related technical reports.
• PyBaMM Documentation: Read the official PyBaMM docs for model options, examples, and API setup.
• NREL Battery Research: Search the National Renewable Energy Laboratory site for battery degradation and EV storage reports.
• Journal of Power Sources: Search the journal for review articles on lithium-ion aging, duty cycles, and state-of-health modeling.
• MIT OpenCourseWare: Look for free course materials on electrochemistry, energy storage, and numerical modeling.
• PubMed: Search for review articles on lithium-ion battery degradation mechanisms and cycle aging.