Identifying and Rank-Ordering Large Volume Leaks in the Underground Natural Gas Distribution System of Massachusetts
Magavi, Zeyneb Pervane
MetadataShow full item record
AbstractRising methane in earth’s atmosphere accelerates climate change, and natural gas is a major driver of that increase. In 2016, Massachusetts passed a law requiring that ‘environmentally significant’ leaks in the aging natural gas distribution network be prioritized for repair in an effort to reduce methane emissions. This research seeks to inform implementation of this legislation by determining an efficient method for identifying and rank-ordering leaks according to volume. Participating gas utilities selected 72 Grade 3 gas leaks across the state, representing all leak-prone pipe materials (cast iron, wrought iron, bare steel, coated steel, and old plastic) and pressures ranging from 0.5 – 99 PSI. A previous study found a heavy-tailed distribution of natural gas leaks by volume in a cast iron low-pressure leak population, and I used the same verified chamber measure of leak volume to test whether that relationship held in a diverse set of pipe materials and pressures. The chamber method flux measures were then used to benchmark other methods, including the initially proposed bar hole method. Suspecting that the bar hole method may be insufficient to rank-order leaks by volume, we also tested four additional methods, including the leak footprint, ringdown spectrometer, FLUXBar , and MSS camera. Both the FLUXBar and the MSS camera are innovative new technologies, with the FLUXBar a co-creation of my research coalition.
I helped to assemble and maintain an unusually broad coalition of stakeholders in order to carry out this research collaboration between the utilities and local environmental organizations, including legislators and regulators to ensure that the science would drive policy. By replicating the heavy-tail distribution in our more varied leak population, I was able to confirm that the identification of the largest volume leaks (LVLs) would maximize emissions reduction per dollar spent. As the ‘bar hole %gas’ method showed no correlation with leak volume, employing it to identify those LVLs would be as efficient as random selection.
Alternate methods did better, with the ringdown spectrometer method, the FLUXBar, and leak footprint methods all correlating with leak volume. The ringdown spectrometer demonstrated the weakest correlation and the use of the FLUXBar is limited by the need for a compressor truck. Leak footprint is not, and can be easily deployed in standard leak surveys. The leak footprint method correlates strongly, so selects leaks that are larger than average leaks, resulting in a return on investment to the ratepayer of less than a year. The originally proposed bar hole method’s return on investment to the ratepayer would be more than six years. These data suggest that the leak footprint is the best available rapid indicator of leak volume, with the FLUXBar measure a potential quantification and verification tool. Continued research is needed as technology-based options advance to meet this challenge.
The stakeholder coalition, in response to these data, agreed to a plan to implement the leak footprint method, accelerate repair, and establish transparency, verification and reassessment processes. The agreement between the three largest gas utilities in Massachusetts and my research team, if enacted as designed over the next two years, has the potential to result in a methane emissions reduction equivalent to 4% of the greenhouse gas emissions inventory for Massachusetts.
Citable link to this pagehttp://nrs.harvard.edu/urn-3:HUL.InstRepos:37945149