Adjusting the Cambridge Bitcoin Electricity Consumption Index for US Energy Prices

Source	Date	Latest Reading
Cambridge Bitcoin Electricity Consumption Index (CBECI)	2023-01-25	115.6 TWh per year
CBECI Adjusted for Electricity Prices	2023-01-25	111.0 TWh per year

---

Last data updates:

• Cambridge Bitcoin Electricity Consumption Index: Sep 3, 2022
• US Energy Information Administration: Sep 3, 2022 (data to May 2023)

---

The Cambridge Bitcoin Electricity Consumption Index is the gold standard of models for tracking the power utilization of the Bitcoin Network, but it is not without flaws. One of these flaws is the model's assumption of constant global electricity prices ($0.05 per kWh). In the current energy shortage, the world is faced with rapidly rising electricity prices and the model must be adjusted.

The attached Jupyter Notebook leverages data available from the US Energy Information Administration and the availability of CBECI data for various energy price assumptions to model the expected network power usage over time for the given price data.

The data finds that although average US industrial electricity prices discounted by 27% to best-match the historic $0.05 per kWh assumption from Cambridge, has historically lined up well with the data, the values in 2022 differ dramatically, suggesting that the CBECI has at times overestimated the electricity demand that the network can support by as much as 20%.

Usage

If the latest dataset is stale, you should update both the CBECI and EIA data.

To update the CBECI data:

• Run the Jupyter Notebook titled generate-cambridge-data.ipynb
• The data will be automatically downloaded and saved into the datasets directory.

To update the EIA data:

• Go to the following link to access EIA data
• Scroll down and ensure that the slider near the top of the table has the latest date selected
• Hit the DOWNLOAD dropdown and select Table (CSV)
• Copy the downloaded .csv file into the datasets folder replacing the stale dataset previously contained.
• Run the energy-adjusted-cbeci.ipynb or test scripts to generate new figures

Limitations

• The CBECI data is only available at $0.01 increments, therefore energy price data in between these increments had to be linearly interpolated.
• The US Energy Information Administration data postings are generally a couple months delayed, so for recent estimates this script forward fills energy prices (assumes them to have not changed since the last data point), which makes the last couple months of data unstable and may also lose seasonal variations.