Analysis and Prediction on Electricity Shortage in South Africa

A team project @ UC x Imagine Scholar Datathon 2020. This project won the Best Project and Most Creative Project.

Context: What is load shedding? Why does South Africa use it?

The energy crisis in South Africa is particularly apparent in the frequency of “load shedding”: a technique of deliberately inducing rolling blackouts nationwide in order to prevent a destabilization and wipeout of the entire national grid. Power is shut off for up to seven hours a day almost every single day. Levels of load shedding range from 1-8. Level 0 means no load shedding.

Content

load_shedding.csv

This dataset gives the stage, date and time of nationwide outages dating back to 2015. The years 2016 and 2017 are excluded because there was no load shedding during those years. (If there is no data for a given day, that means there is no load shedding on that day. Or you can say, it is level 0 load shedding)

Total_electricity_production.csv

This dataset gives the electricity produced in South Africa by gigawatt hours, going back to 1985. It is measured quarterly.

Electricity_indicators.csv

The datasets includes different electricity-related indicators of countries including South Africa, World Average, Nigeria and etc.

Acknowledgements

Great thanks for the Imagine Scholar x UC DS Communities to organize this event!

Inspiration

1. How is South Africa's electricity access rate compared to other countries with similar GDP?
2. Is the South Africa's electricity access rates different between rural and urban populations?
3. Are there any seasonality in South Africa's electricity production? Can you information from other place to explain the seasonality?
4. Is the frequency of nation wide load shedding increasing or decreasing?
5. Can you predict electricity production for next year (2021)>
6. Can you predict the number of nation wide outages for next year?