Determine method for county-level emissions

[calmc]

Intro

There is an audience for and update to the Industry Energy Databook (IEDB) all-industry energy data estimates at the county-level. The IEDB used establishment counts by NAICS code, employment size class, and county from Census County Business Patterns data, together with energy intensity values calculated from Manufacturing Energy Consumption Survey (MECS) data (see the code), for part of its energy estimation process for facilities that do not have large enough GHG emissions to report to the EPA's GHGRP. The employment size class was used to differentiate the energy intensities of facilities that are different sizes. This differentiation follows MECS data from Table 6.4.

Problem

Disclosure rules were changed between the 2016 and 2017 CBP releases and now most of the establishment counts are withheld. So, the same process used by the IEDB to estimate county-level energy use can't be used now. There needs to be a new way of estimating energy use for facilities that would've been covered by this approach.

Proposed Fixes

1. Use the 2017 FIED's underlying FRS data as establishment counts for manufacturing, construction, and mining sectors

• Check the FIED/FRS counts by county against 2016 CBP data. See how close they are and if there are any gaps
• However, this doesn't fix the use of employment size classes to adjust energy intensity estimates from MECS.

2. Estimate manufacturing energy intensities

[calmc]

Another proposed fix:
Start with EIA estimates of industry energy use by state (EIA SEDS) and allocate to facilities that aren't captured by the approaches that use NEI and GHGRP data to derive energy estimates.

[calmc]

Another thought: it might be helpful to see Issue #9 and better understand which facilities are missing energy estimates (or have energy estimates = 0) AND are NEI-reporters (i.e., have an EIS ID).