GridPIQ reports fossil fuel costs as a project impact. It is important to note that these costs are based on the price of fuel, and don’t encapsulate other operational or capital costs involved in electricity generation. In order to estimate these fossil fuel costs, state level cost data ($/million BTU) by fuel (coal, natural gas, petroleum) are obtained from the U.S. Energy Information Administration (EIA) through their application programming interface (API). The specific data set can be found here. GridPIQ uses annual data for each state for the “Coal,” “Distillate Fuel Oil,” and “Natural Gas including Supplemental Gaseous Fuels” categories. In the case of missing data, the previous year’s data is used.
In order to map annual state level costs to AVERT regions, electric power data has been obtained from the EIA’s Form 923. For each year, electric power generation data for all electric generators with a nameplate capacitor of 1 MW or greater is used. Coal, natural gas, and petroleum products are tracked separately.
Generation data from individual plants is mapped to each of the AVERT regions using data from the EPA’s Emissions and Generation Resource Integrated Database (eGRID). The result of this plant to AVERT region mapping is a mapping of state contributions to AVERT region for each year by fuel.
With a state to AVERT mapping in hand, the state fossil fuel costs ($/million BTU) from the EIA are aggregated to the AVERT region level. Next, AVERT’s statistically aggregated and binned heat-rate data (million BTU) for each fuel is multiplied by the estimated AVERT fossil fuel costs. The result is approximate fossil fuel costs for a set of system power levels. After pre- and post-project load profiles have been scaled and run through fossil fuel correction, linear interpolation is used to compute costs for each hour given a system power level. This interpolation procedure is identical to how GridPIQ computes emissions from the AVERT data.
NOTE: At the time of writing (01-26-2018), the EIA did not have fossil fuel costs for 2016. Costs for 2016 are estimated by simply using costs from 2015.