Colorado State University researchers release comprehensive study of methane emissions from natural gas gathering and processing facilities — link updated

FORT COLLINS —In the largest, most comprehensive study ever conducted on methane emissions from natural gas gathering facilities and processing plants, researchers led by Colorado State University found that 0.47 percent of the methane produced domestically is lost during gathering and processing operations. According to the study, methane emissions from gathering systems are equivalent to 30 percent of overall methane emissions in the current U.S. greenhouse gas inventory. The majority of these methane emissions were attributed to normal operations of gathering facilities.

The study, published today in the journal Environmental Science & Technology, was based on onsite measurements conducted at 114 natural gas gathering facilities and 16 processing plants in 13 states over 20 weeks. It is part of a series of methane emissions studies organized by Environmental Defense Fund, representing the largest measurement campaign of the U.S. natural gas infrastructure to date.

Gathering facilities collect and compress natural gas from multiple wells then send the gas to a transmission line or a processing plant where ethane and natural gas liquids are removed. In addition to large natural gas-powered compressors, both gathering and processing facilities include equipment such as liquid separators, dehydration systems, acid gas removal systems and liquid storage tanks. Processing plants also include additional systems to remove the ethane and/or liquids.

While all processing plants are required to report methane emissions under EPA’s Greenhouse Gas Reporting Program, very few gathering facilities are currently required to report methane emissions.

“This was the first-ever national study on methane emissions from natural gas gathering facilities,” said Anthony Marchese, professor of mechanical engineering at CSU and principal investigator. “We could find only two prior measurements of gathering facilities in the literature, and no national inventory of gathering facility counts or emissions from gathering facilities exists.”

Researchers found that methane emissions from gathering facilities are substantial – 1,697 Gigagrams (Gg) per year — while emissions from processing plants are less than half that amount — 505 Gg per year. These emissions are equivalent to the amount of natural gas consumed annually by 3.2 million U.S. homes and represent a potential loss of $390 million in revenue to producers. 

The 2014 EPA Greenhouse Gas Inventory estimates overall methane emissions for the U.S. natural gas supply chain at 6,186 Gg.

The primary component of natural gas, methane is a greenhouse gas many times more potent than carbon dioxide when released into the atmosphere unburned. The nation’s vast natural gas infrastructure – including wells, compressor stations, processing plants, pipelines, and storage facilities – constitutes the largest industrial source of methane emissions in the United States.

The researchers also identified opportunities to improve the two federal programs that track methane emissions from the natural gas supply chain, EPA Greenhouse Gas Reporting Program (GHGRP) and EPA Greenhouse Gas Inventory (GHGI). The GHGI does not explicitly break out gathering facilities as a category separate from field production. Marchese and his team, which also included researchers from Carnegie Mellon University and Aerodyne Research, analyzed the current production inventory to determine the appropriate fraction of emissions to assign to gathering, and the results were eight times higher than the best interpretation of EPA data for gathering facilities.

“The results for processing plants are lower than the EPA estimate in part because we observed fewer high-emitting reciprocating compressors in actual use than the number assumed in the EPA Greenhouse Gas Inventory,” Marchese said. “Only one of the sampled processing plants had a methane loss rate greater than 0.5 percent, whereas 40 percent of the gathering facilities had methane loss rates greater than 0.5 percent. We believe that this could be a consequence of differences in regulations and reporting requirements between the two types of facilities.”  

Processing plant methane emissions are reported to the GHGRP. Depending on the type, many processing plants are also subject to regulations that require companies to periodically monitor their facilities for leaks and make a first attempt to repair any leaks found within five days. Gathering facilities are not presently subject to those regulations, Marchese explained, and only a small fraction of gathering facilities report methane emissions to the GHGRP.

The study also compared processing plant emissions to those in the GHGRP and found methane emissions from processing plants were three times higher than the GHGRP.

Currently, methane emissions reported from gathering facilities are very low — less than 1 Gg per year – because very few of these facilities are required to report methane emissions according to current EPA rules.  The EPA has proposed a rule that would require gathering systems to report methane emissions.

The CSU team did not measure emissions from natural gas gathering lines in this study. Future studies need to be conducted to accurately determine the total amount of gathering pipelines in the U.S. and their methane leakage rates.

This work was financially supported by the Environmental Defense Fund (EDF), Access Midstream, Anadarko Petroleum Corporation, Hess Corporation, Southwestern Energy and Williams Companies. DCP Midstream, while not a financial sponsor, provided site access for one of their processing plants selected at random.

The Colorado State University study is one of 16 organized by the Environmental Defense Fund and industry partners to better quantify the amount of methane released into the atmosphere from the natural gas supply chain. This is the third paper published by researchers at CSU, Carnegie Mellon and Aerodyne Research as part of the project.

The full paper can be found at http://pubs.acs.org/doi/abs/10.1021/acs.est.5b02275

-30-