Pie Chart

 When a Pie Chart Actually Makes Sense: U.S. Electricity 

 A strong use case

For this project, I visualized the U.S. utility-scale electricity generation mix in 2023 using a pie chart. The dataset sums to a meaningful whole (100%) and contains five clear categories: natural gas, nuclear, coal, renewables (total), and petroleum & other. That setup fits the textbook criteria for a pie chart: showing parts of a whole, a small number of categories, and differences that are visually distinct enough to compare.

https://www.eia.gov/energyexplained/electricity/electricity-in-the-us-generation-capacity-and-sales.php?utm

The audience’s task is to grasp the concept at a glance, not to make ultra-precise comparisons between similar values. Visualization research generally finds bars are more accurate for exact comparisons, but pies are acceptable when the primary question is “what share does each piece contribute to the whole?” and the slice count is limited.

Data (EIA, 2023): Natural gas 43.1%, Nuclear 18.6%, Coal 16.2%, Renewables (total) 21.4%, Petroleum & other 0.8%. These categories are mutually exclusive and sum to 100%, which is essential for a valid pie.

https://www.eia.gov/energyexplained/electricity/electricity-in-the-us-generation-capacity-and-sales.php?utm

I built a flat (non-3D) pie chart with direct data labels showing both the category and percent, a descriptive title, and an in-chart source note. The focus is on clarity: no exploding slices, no gradients, no shadows. Avoiding 3D prevents distortion of perceived angles and areas.

If the goal were precise comparison between similar slices, a bar chart would support more accurate judgments for a viewer because length is read more reliably than angle or area. If I needed to emphasize the total while also leaving room for annotations in the middle, a donut chart would be acceptable as a stylistic variant of a pie.