Data Center CO₂ Projections

Interactive projection

Select a state and a year. The model evaluates both fits at the chosen year and returns their product.

Projected total emissions

Shown on a log scale. Dashed lines show the fitted model; markers show observed MW multiplied by a hybrid emission rate.

MW capacity

Data-center nameplate capacity by state.

Emission intensity

Grid CO₂ output rate, in pounds per MWh.

Methodology

Two quantities drive total emissions: installed data-center capacity and the emission intensity of the grid supplying that capacity. Both show roughly exponential trends over the available data, so we fit each separately and take their product.

1. Capacity fit

MW nameplate data for 2021 through 2024, together with a published 2030 industry projection, is fit to the model:

MW(t) = a · e^b(t−2021)

The fit is performed with SciPy's curve_fit, which applies nonlinear least-squares minimization. Anchoring the exponent at 2021 means that the coefficient a equals the 2021 baseline capacity, which simplifies interpretation of the returned parameters.

2. Emission-intensity fit

The same functional form is used for emission intensity, with the anchor shifted to match the data window. EPA eGRID state totals for 2018 through 2023 are fit to:

E(t) = a · e^b(t−2018)

The initial guess supplied to curve_fit must use a negative value for b, for example p0 = (E₂₀₁₈, −0.1). With a non-negative seed the solver can converge to a nearly flat fit for states in which the grid is decarbonizing quickly. Illinois exhibits this behavior most clearly.

3. Annualizing

MW is a rate rather than an energy quantity. To convert to annual energy we multiply by 8,760, the number of hours in a non-leap year. Because data-center load is approximately continuous, this factor is appropriate as a first-order estimate:

MWh(t) = MW(t) · 8760

4. Total emissions

The total CO₂ in a given year is the product of annual energy and the corresponding emission intensity:

Total CO₂(t) = a_MW · e^{b_MW(t−2021)} · 8760 · a_E · e^{b_E(t−2018)}

Since both factors are exponential, the product is exponential as well, with growth rate equal to the sum of the two component rates. Virginia's combined rate is large and positive. In the other four states, the decline in emission intensity partially offsets the growth in capacity.

5. Fitted parameters

The coefficients returned by curve_fit for each state are shown below.

State	a_MW	b_MW	a_E	b_E

6. Caveats

We only have five years of MW data and six years of emission data per state, so the fitted growth rate has a wide margin of error.
The 2030 MW number comes from an industry forecast, not a real measurement, which means it has an outsized effect on where the projection lands.