A study on the consumption of copper to scale with the growth of electric vehicle implementation by 2030
Cu on the Otherside
-
[This project aims at ]
- [Scenario]
- Goals
- Project Dependencies
-
- Scope
- Acquiring
- Preparing
- Data Dictionary
- Key Findings
-
- Hypothesis
- Steps
- Conclusion
The transition to full electic vehicles is well underway in the United States. Climate change is at the forefront of politics. President Biden proposed a climate change initiative aiming for 50% of all vehicles sold in the United States to be electric by 2030. In addition to 50% electric vehicle sales, President Biden also propsed 500,000 additional chargers to accomodate the amount of new EVs on the market. An electric vehicle uses almost four times the amount of copper than a regular car at almost 200 pounds per passenger car, while commericals vehicles can use up to 800 pounds. Our goal is to use time forecasting to predict the total amount of new us vehicle sales sold in 2030, total refined copper produced both in the US and world, to see if this the proposal is feasible.
- First we will use machine learning and time forecasting to predict total New US car sales in 2030. Using the predicted forecast, we will calcuate the proposed 50% sale goal and use that goal to determine the amount of copper needed to meet that goal.
- Second we again use machine learning and time forecasting to predict the yearly total mined refined copper produced both domestically and globally. Utilizing the forecast and the current rate of copper allocated for transportation of 16% to determine feasibility of the proposal.
- Additionally, we extrapolated the 500,000 proposed charging stations out to 2030 to assess the needed copper to meet the demands of the proposals.
- Showcase the results of our investivations in a presentation delivered to stakeholders.
- Create a reproducible jupyter notebook report that includes process, takeaways, and discoveries from every stage of the multi-pipelines
- A whitepaper to showcase the research and findings from our investigations.
- Please find the deliverables here:
- ** We have included both the cleaned and prepped data CSVs as well as the unprepped. For easier reproducability, please use the cleaned and prepped data.
- Read and follow this README.md
-
- Finally run our *FINAL NOTEBOOK replicate our models and analysis
This project makes use of several technologies that will need to be installed
Dependencies can be installed quickly with just a few lines of code.
%pip install notebook
%pip install numpy
%pip install pandas
%pip install matplotlib
%pip install seaborn
%pip install scipy
%pip install sklearn
- Total New Car Sales: https://www.bts.gov/content/annual-us-motor-vehicle-production-and-factory-wholesale-sales-thousands-units
- US Copper mining production/consumption: https://www.usgs.gov/centers/national-minerals-information-center/copper-statistics-and-information
- To extract the data from the PDF we used Tabula
- Historic Prices of Copper - https://www.macrotrends.net/1476/copper-prices-historical-chart-data
{ How many records/columns? How many nulls? Does this project focus on a particular subset of the overall data? }
- The data was acquired from the "About the Data" section
- Each dataset was prepared in individual notebooks. Please see the perspective notebook on how hte data was prepared.
- If the raw data is used: follow the functions in prespective notebooks to have the data split and ready to explore for analysis.
- Ran vizulations to determine seasonality, trands, and how each target variable changed over time.
- Ran statistical tests to determine stationality of the data and the correlation of lag
- Data was processed during the prep stage for exploration
- Developed a baseline by taking a simple moving average for each target variable
- Due to the small amount of available data, we split the data only into train and test. This gave us the most amount of data to train the models on.
- Built multiple forecasting models to improve upon baseline accuracy.
- A final reproducable jupyter notebook
- Slideshow to show the important findings during the investigation
- Whitepaper to detail findings and projections in more detail
| Variable name | Explanation | Unit |
|---|---|---|
| year | Year of the observation | DateTime |
| prod_mine | Gross weight of unrefined material mined | 1K metric tons |
| refine_from_ore | Weight refined from ore | 1K metric tons |
| refine_new_scrap | Weight refined from scrap generated by production process | 1K metric tons |
| refine_old_scrap | Weight refined from recycled consumer products | 1K metric tons |
| import_ore | Weight of ore imported to the united states | 1K metric tons |
| import_refined | Weight of refined copper imported to the united states | 1K metric tons |
| export_ore | Weight of ore exported from the united states | 1K metric tons |
| export_refined | Weight of refiend copper exported from the united states | 1K metric tons |
| consumption_refined | Weight of refined copper used by the united states | 1K metric tons |
| consumption_scrap | Weight of copper scrap used by the united states | 1K metric tons |
| total_consumption | Weight of all copper used by united states | 1K metric tons |
| total_production | Weight of all copper produced by the united states | 1K metric tons |
| copper_for_cars | Weight of copper allocated for transportation | 1K metric tons |
| Variable name | Explanation | Unit |
|---|---|---|
| year | Year of the observation | DateTime |
| mine_production | Gross weight of unrefined material mined globally | 1K metric tons |
| refined_production | Total weight of global copper production | 1K metric tons |
| refined_usage | Total weight of copper used globally | 1K metric tons |
| Variable name | Explanation | Unit |
|---|---|---|
| year | Year of the observation | DateTime |
| total_sale | Number of cars sold in the united states | Thousands |
- 5,033,500 cars to be electrified
- Deficit of 312k Metric Tons
- Imports included in US Production
- Models do not account for the increase in copper for other industries
- CO2 savings make up for the increased emissions of mining
- The US lacks sufficient production. New mines will not come online soon enough to be meet the goals needs.
- Pay more per ton for imports. However, this would increase the cost of vehicles as well.
- Increase Global Prod. ~1.5% w/ no other demand. No other country could import any of the increased global
- Another way to increase supply is to increase recycling can help make up deficits. Copper is infinitely recyclable.
- Propose a plan to go 50% hybrid cars instead of full electric
- Short 21k metric tons instead of 312k metric tons. More possible to make up this difference in imports.
- Overall savings of 6,616,833 CO2 emissions after taking mining CO2 into effect
- No additional infrastructure needed such as charging stations
- Continue to build the copper price model
- Look at global sales and see how that affects global copper supply