Emissions of methane, nitrous oxides, nitrogen oxide, and sulfur dioxide, are based on data provided by the Environmental Protection Agency (EPA) in the eGrid 2014 data file in the 2014 Emissions and Generation Resource Integrated Database (eGrid). Emissions are given in units of pounds (for methane and nitrous oxide) and tons (for other gases) emitted into the air as a result of all electricity generated in a particular state on an annual basis.

Emissions of greenhouse gases are also based on data provided by the Environmental Protection Agency (EPA) in the eGrid 2014 data file in the 2014 Emissions and Generation Resource Integrated Database (eGrid). Emissions are given in units of tons of carbon dioxide equivalent gas emitted into the atmosphere as a result of all electricity generated in a particular state on an annual basis. Greenhouse gases that are not carbon dioxide are converted to carbon dioxide equivalent emissions by multiplying their greenhouse warming potentials. For example, methane gas has a greater greenhouse warming potential than carbon dioxide — about 25 times more. Therefore, one ton of methane is about the same as 25 tons of carbon dioxide equivalent gas. A useful greenhouse gas equivalency calculator is provided by the EPA here.

Estimates of mercury, lead, and arsenic emissions were made based on EPA NEI 2014 (National Emissions Inventory) data for total emissions from fuel combustion for electricity generation and total electricity generation from coal based on EPA eGrid 2014 data. For example, a state's emission of arsenic resulting from generating electricity was estimated by first dividing the total pounds of arsenic emitted in the U.S. as a result of electricity generation (EPA NEI data) by the total MWh of electricity generated from coal burning in the U.S., to arrive at an emissions factor in pounds/MWh. This emissions factor was then multiplied by each individual state's electricity generation from coal to provide an estimate of total arsenic (in pounds) generated by that state every year as a result of generating electricity. These calculations assume that almost all arsenic, lead, and mercury emissions are a result of burning coal and do not take into account significant variations among coal-based power plants resulting from various clean coal technologies installed at a particular plant. Clean coal technologies can reduce emissions by 90% or more.

Water consumption resulting from generating electricity is calculated first for all electricity generation methods (except hydroelectric power), using calculated estimates for power plants in the U.S. from the Union of Concerned Scientists (UCS) Energy-Water database. Estimates for water consumption from hydroelectric power are made using hydroelectric site water consumption factors derived by the National Renewable Energy Laboratory (NREL). Estimates for water consumption by hydroelectric power generation are complex because not all hydroelectric power is generated from manmade reservoirs (i.e. from natural lakes, the water would evaporate regardless of whether hydroelectric power was being generated or not) and because manmade reservoirs have multiple uses (e.g. power, recreation) that make quantifying water consumption for a single use difficult. The numbers provided in this analysis assume all evaporated water from reservoirs associated with hydroelectric facilities are a result of electricity production and therefore represent maximum values.

• United States (EPA) Environmental Protection Agency. (2014). Emissions & Generation Resource Integrated Database.
• United States (EPA) Environmental Protection Agency. (2014). National Emissions Inventory.
• National Renewable Energy Laboratory (2003). Consumptive Water Use for U.S. Power Production.