Introduction

This analysis uses the Energy Usage Analysis System dataset, which tracks energy consumption across government facilities. It focuses on electricity, natural gas, fuel oil, chilled water, and steam usage from 1973 to 2019, as published by the U.S. General Services Administration (https://catalog.data.gov/dataset/energy-usage-analysis-system ).

After extensive data cleaning and transformation, I focused on the top 12 facilities by total energy usage. Rows with null values, negative or zero energy usage, and extreme outliers (e.g., values exceeding 1e14 kBtu) were removed to ensure meaningful comparisons. All selected facilities have total usage in the trillions of kBtu or less, providing a consistent and interpretable dataset for visualization.

Energy Visualization Tabs

Facility Locations and Size

This map highlights the city and state of each facility. The size of each circle corresponds to the facility’s gross square footage, with larger circles representing bigger facilities.

Total Energy Usage by Facility

The total usage of each energy source from 1973 to 2019 was calculated for each facility, and the facilities were then ordered from highest to lowest total energy consumption.

# SECOND VISUALIZATION
# lollipop chart, total usage per name in top twelve 

palette_colors <- paletteer::paletteer_d("palettetown::smoochum")
selected_color_x <- palette_colors[c(5)]
selected_color_y <- palette_colors[c(12)]
top12_facility_usage$NAME <- tools::toTitleCase(tolower(top12_facility_usage$NAME))

ggplot(top12_facility_usage, aes(x = reorder(NAME, TotalUsage), y = TotalUsage / 1e12)) +
  geom_segment(aes(x = NAME, xend = NAME, y = 0, yend = TotalUsage / 1e12), color = selected_color_x) +
  geom_point(color = selected_color_y, size = 4, alpha = 0.6) +
  theme_light() +
  coord_flip() +
  theme(
    panel.grid.major.y = element_blank(),
    panel.border = element_blank(),
    axis.ticks.y = element_blank()
  ) +
  labs(title = "Total Energy Usage by Facility ", y = "Total Usage (T kBtu)", x = "")

Usage by Energy Type

This stacked bar chart displays the energy mix of each facility, ordered from highest to lowest total energy consumption. Hover over the colored segments to see exact values for each energy source.

Proportion of Facility Usage

This first pie chart compares the total energy consumption of each facility, showing which buildings use the most energy.

Proportion of Energy Type

This pie chart breaks down the overall energy consumption by type, illustrating the proportion of different energy sources used across all facilities.

Electricity Usage

This is a stacked area chart showing how much electricity each facility used from 1973 - 2019.

top12_electricity$NAME <- tools::toTitleCase(tolower(top12_electricity$NAME))
ggplot(top12_electricity, aes(x = FYR, y = Total_Electricity, fill = NAME)) +
  geom_area() +
  labs(title = "Stacked Electricity Usage Across Facilities",
       x = "Year", y = "Total Electricity (kWh)") +
  theme_clean() +
  theme(plot.title = element_text(hjust = 0.5)) +
  scale_fill_manual(values = palette_colors) +
  scale_x_continuous(breaks = seq(min(top12_electricity$FYR), max(top12_electricity$FYR), by = 5), minor_breaks = seq(min(top12_electricity$FYR), max(top12_electricity$FYR), by = 1), labels = scales::label_number(accuracy = 1, big.mark = "")) +
  scale_y_continuous(
    breaks = pretty_breaks(n = 12),
    labels = comma
  )

Energy Efficiency

This scatterplot illustrates the energy efficiency of each facility. Energy Usage Intensity (EUI) is calculated by dividing a building’s annual total energy consumption (in kBtu) by its gross square footage. A lower EUI indicates that the building is using energy more efficiently.

Conclusion

My analysis and these visualizations provide a broad overview of the energy consumption patterns across a subset of government buildings. One notable finding is that fuel oil represents the largest energy source. It also appears that larger buildings tend to exhibit greater energy efficiency compared to smaller ones; a logical next step would be to compare the EUI of buildings within specific categories. While I would have liked to explore renewable energy usage and visualize trends over time, missing data limited this analysis. Nevertheless, the dataset and these visualizations offer a clear and informative summary of energy consumption for these buildings.