qlangfitt
commented
3 years ago Looks good. Storage wells are kind of strange because clearly it is not referencing the number of storage wells on an oil and gas field, but rather taking into account losses from them further down the natural gas supply chain. Therefore, seems like if any factor is used it should just be generic and apply equally to all exported gas. I can think of at least two options, one simple and one complicated.
Simple option: Use the total storage wells emissions from the US EPA GHGI for 2018 (15.4 kt CH4) and divide by total US production to get a fractional leakage rate. Then just apply that fractional leakage rate to all exported gas. This would be like a population factor and we wouldn't have to guess what fraction of gas goes into storage or not (although maybe we're already accounting for that?).
Complicated option: Perhaps, if DrillingInfo has the "production" data for all US storage wells, we could just run those through OPGEE's component fugitives tools as if they were normal wellheads, get the gas-weighted average loss rate, and apply that loss rate to all gas that passes through storage (would have to assume a % of gas going through storage). My two cents is that this option is probably a lot of work, complexity, and assumptions for something that's likely to be a pretty small portion of total emissions.
JSRuthe
arbrandt
This issue combines issues #257 and #259. These issues both reference fugitive streams associated with offsite (off of well-pad) transmission and storage equipment.
Some time ago, @arbrandt and I elected to handle all off-site fugitive emissions using the site-level approach. Therefore, the user has the option of using the component-level model or the site-level model for on-site equipment, but only the site-level model can be used for off-site equipment (which includes gathering and boosting, transmissions and storage, and distribution infrastructure.
Loss-rate estimates from the Zimmerle et al. (2015) paper are applied to venting and fugitive emissions streams for transmission and storage. Zimmerle et al. is a modelling study fitted using data from Subramanian et al. (2015), who make measurements at 36 transmission compressor stations and 9 storage compressor stations. Unfortunately, the paper is unclear on exactly which equipment are measured (many are bundled within a category of "non-compressor component fugitives"). I've emailed Dan Zimmerle and learned that "non-compressor equipment" includes storage separators but does not include storage wellheads.
Based on this information, several emissions streams in OPGEE will need to be eliminated to avoid double counting with the transmission and storage loss rates (as these loss rates contain emissions from multiple equipment). These emissions streams are summarized in the Table below:
@arbrandt and @qlangfitt please comment on this approach. I'm not sure exactly how to handle storage wellheads as it is difficult to justify how many to allocate to single field. As a reference, there are ~3300 storage wells in the US (based on DrillingInfo), which seems very small compared to the total well count.