Families on the front lines of mining, drilling, and fracking need your help. Support them now!


Last week Earthworks published an analysis comparing Environmental Defense Fund flyovers of flares against the Texas Railroad Commission’s flaring permit database for those flares observed on Texas General Land Office lands. The results: 75% of observed flares were unpermitted.

In a story about our analysis, Railroad Commission spokesperson Andrew Keese responded:

“The RRC’s commitment to further reducing flaring is reflected in actions commissioners took last November to revamp the application for exceptions. The most recent RRC production data shows that in the last year and half, the percentage of natural gas flared compared to the natural gas produced in Texas dropped from a high of 2.3% in June 2019, to 0.77% in November 2020. During the same period, the volume of gas flared decreased by more than 71%.”

Let’s parse this response. First, the Railroad Commission does not deny or question our central finding — 75% of flares are unpermitted — in any sense. This suggests either our analysis is sound or the Railroad Commission doesn’t have the data to disprove it.

This first point means the rest of Mr Keese’s response is valueless. If 75% of flares are unpermitted, they’re unreported (because otherwise the RRC would require them to get a permit). If, as seems likely, Mr Keese’s response is based on what could be as little as ¼ of flares that bother to get a permit then the Railroad Commission’s data on flaring vastly undercounts the total number of flares, and those that it does count are likely unrepresentatively “responsible”. If the RRC doesn’t bother to check if operations get flaring permits and instead relies on operators to inform the RRC of said flares, it’s likely that those operators care more about compliance than those that don’t self report.

To reiterate, our analysis suggests that 75% of flares in Texas are unpermitted. If flares aren’t permitted the RRC would not include them in their flaring data, meaning the data they presented in their response to our analysis, as well in most of their public discourse on flaring, severely underrepresents actual flaring.

Feb 04 Update:
After this blog was published Andrew Keese, the RRC spokesperson referenced in the original post, reached out to Earthworks to clarify his statement as well as to pass along the full text. Rather than denying the analysis presented by Earthworks, the RRC’s statement seeks to explain why 75% of flares would be allowed to go unreported and unpermitted. 

The full statement also includes: 

“that rules specify certain situations in which an operator can flare, including for safety reasons, without going through the application process to obtain an exception from the RRC.”

While this is true, those guidelines only include a few specific instances of allowed flaring such as during maintenance activities, in the 10 days after a facility first begins operating or if the flaring is incredibly low volume. 

The data we’ve collected shows that 30 of the 34 unreported flares documented by EDF were documented flaring during at least two of the flyovers, which occurred months apart. It is unlikely that those sites had two separate non-reportable flaring events just a few months apart. If the reason the flares documented by EDF were unreported was due to low volume, that alone should be cause for concern by the RRC. It is possible that the flares observed may be low volume on an individual basis but if 75% of flares are unreported and unpermitted due to volume that much flaring in the aggregate would significantly alter flaring data published by the RRC. 

However, it seems unlikely that flares such as Diamondback Longfellow 3-13 (which despite being documented flaring three times by EDF and twice by Earthworks has not reported a flare since 2018) would be considered low volume: 

The upshot: the Railroad Commission confirmed our analysis by trying to excuse it.