California’s Energy Systems

Project Summary

I analyzed data in Google Colab (Python) from the previous 20 years (2001 to 2021) of California's energy reports to address the following questions in the wake of the rolling blackouts in the state over the summer of 2022:

• What does this mean for the future of California’s energy grid?

• As California attempts to phase out fossil fuels, how will it provide electricity to all its citizens?

I processed the data and used various dimension reduction techniques such as principal component analysis (PCA) and multi-dimensional scaling (MDS). I created several visualizations—line graph, stacked bar charts, scatter plot matrix—and interpreted the results. I created several data mining models including simple linear regression, multiple linear regression, LASSO regression, k-NN regression, and k-NN classification. I compared the performance evaluation metrics of each model to find the best model. I presented my work in a 30-page report and a seven-minute presentation.

Project Links

CA Energy Systems: Revisited

After a couple of years, I revisited this project. I used Tableau to visualize the data and create two dashboards, shown below and linked here and here. The following is a brief overview of the visualizations.

The first interactive dashboard shows the total capacity (MWh), prime fuel use (MMBtu), and net MWh generated for all plants, color-coded by energy type. These graphs emphasize the significant amount of energy produced by natural gas. Certainly, the increase in both capacity and net megawatt-hours of generation by solar energy in 2012 does not go unnoticed.

The second dashboard, below, shows the average capacity (MWh), prime fuel use (MMBtu), and net megawatt-hours generated for each plant color-coded by energy type. Notice, as mentioned in my original report, that nuclear energy has had the greatest average capacity, average prime fuel use, and, most importantly, average net generation.

Natural gas plants have a low average of net megawatt-hours generated; however, the total megawatt-hours generated by natural gas plants are much greater than any other type of plant. Without running the numbers (though this was done in Python in the original report), this implies that many natural gas plants are serving California. Comparably, the average generation and capacity of far fewer nuclear energy plants is much greater.

These results beg the question of why California would decommission its last nuclear plant. If a nuclear plant averages far greater megawatt-hours of energy, why would California decommission these plants and hold onto many more natural gas plants that generate much less energy?

Next, I will update the data.