The Bubble/Slug Flow Model for Methane Leakage from natural Gas Wells as an Analogue for Shallow CO2 Migration

Abstract

Previous studies of CO2 leakage from wells have assumed that either CO2 leaks into aquifers in the form of a radial lateral buoyant lateral plume of gaseous CO2 or as a lateral radial plume of CO2 dissolved water. These assumptions are not based on any actual observations of the nature of gas leakage from well bores. No direct information appears to exist on CO2 leakage from actual CO2 wells so to understand better this phenomenon. In fact the seminal papers on this topic may have chosen these models because of ease of mathematical manipu lation. A wide range of information on the nature of leakage of methane from natural gas wells has been compiled. Analysis of this evidence suggests that methane neither forms buoyant radial plumes of gas, nor radial plumes of dissolved methane. Rather methane appears to be transported dominantly vertically by a combination of bubble and slug flow. Bubble flow occurs by bubbles of methane (or CO2) gas buoyantly rising up fractures. Slug flow is initiated when multiple bubbles amalgamate in a fracture to form a film that moves as a single mass. Slugs have very low surface area and therefor minimize the dissolution of the gas phase. Both bubbles and slugs in fractures move in an essentially vertical direction. The evidence for bubble and slug flow of methane comes from a variety of observational sources including: 1) measurement of methane concentrations radially away from a leaking well over a period of years; 2) evidence of gas scavenging (a phenomenon that cannot occur if the gas is in the dissolved phase); and 3) observations of concentrated bubbles emerging immediately around a gas wells following a sub -surface blowout while measurements of dissolved methane in surrounding water wells demonstrated minimal methane concentrations.Methane gas has also be directly imaged by down-the-hole cameras, emanating from fractures in the side of a water well as a string of bubbles. In this case again little if any methane dissolution into the aquifer had occurred. As a result damage to vegetation around leaking natural gas wells is typically found to be a radial zone around the well head less than a meter in diameter. An assumption of rapid CO2 dissolution into aquifers assumes that there is intimate contact between CO2 and water or brine. If CO2, like methane is dominantly transported by a combination of bubble and slug flow then the surface area of CO2 exposed to the water phase will be relatively minimal and little dissolution will occur. If CO2 gas leaking from a well into an aquifer, behaves in a similar way to methane many of the negative environmental consequences (such as metal contamination resulting from dissolved CO2 lowering the aquifers pH), will not occur. This analogue, if applicable, has significant implications for how wells should be monitored for leakage. Current monitoring conceptualizations based on monitoring wells are probably not the best approach.