Subsurface methane migration from natural gas distribution pipelines as affected by soil heterogeneity: field scale experimental and numerical study

Abstract

With the increased use of natural gas as a transition fuel on the path toward low-carbon energy, safety and environmental concerns from leaking underground natural gas pipelines are becoming more widespread. Pipeline leakage can be catastrophic when flammable mixtures of methane and air accumulate and ignite. Methane is additionally a powerful greenhouse gas thus; leakage also poses adverse climate impacts. What is not well understood in leakage incidents is how the environmental conditions (i.e. soil type and moisture, competing utilities, ground surface cover) affect gas migration behavior (including the migration rate and extent). To mitigate these concerns, an increased understanding of gas migration through soil is necessary to support efficient leak response, improve leak detection technology, and inform policy and regulations. This study investigates the effects of various soil conditions including soil moisture, heterogeneity, gas leakage rate, and pipeline depth on natural gas migration caused by leaking pipelines. Eight field-scale experiments were performed in custom designed test beds with different soil textures, pipeline burial depths, and alterable leak rates and soil moistures. Subsurface and surface methane concentrations in addition to soil moisture, temperature, and meteorological data were collected over time under transient and steady state conditions. Results found pronounced effects of soil heterogeneities and to a lesser degree of soil moisture on the shape and travel distance of the subsurface methane plume. Leaks from shallow pipelines were found to produce higher surface methane concentrations than from pipelines buried deeper. Advection and buoyancy played a large role on methane concentrations close to the leak, whereas diffusion dominated as the plume moved farther away from the source. Although the conditions explained above were considered separately, they affect each other, yielding a combined and complicated effect on gas distribution profiles. Numerical modeling studies also ensued using TOUGH2 EOS7CA. Comparison to experimental results demonstrated the model’s ability to accurately simulate methane migration from a leaking pipeline. The findings of this study will aid leak detectors locate leaks more effectively and regulators create more knowledge based decisions to increase resident safety and decrease climate forcing emissions.