Abstract
We compare the observed dynamics of subsea starting plumes with multiphase computational
fluid dynamics (CFD) simulations. In the event of subsea loss of containment of either natural gas or carbon
dioxide it is important to understand how the gas disperses in the ocean column and acends to the surface where
it can pose a risk to installations, vessels and on-board personnel. Risks factors include; fire and explosion
hazards, sudden and persistant hydrodynamic loads, in case of carbon dioxide the hazard of asphyxiation, and
in case of a gas rich in hydrogen sulfide toxic effects. Thus, it is routine to conduct atmospheric dispersion
simulations to estimate size of exclusion zones on rigs and safe stand-off distances for vessels in case of loss of
containment. Atmospheric dispersion is today in some respects seen as a routine and mature exsersize. Howerver,
subsea dispersion of gas and surfacing of this gas is still an area with limited data and large uncertainties
with respect to a number of factors (Olsen & Skjetne(2016-A)); transient release rates, depth, mass transfer
to ocean, degassing at ocean surface. A key obstacle to obtaining good data has been the challenge of subsea
imaging and characterization of large transient bubble plumes structures. Here we utilize recent developments
in 3D temporal sonar imaging obtained for starting plumes from 30 meters depth with predictions obtained
using computational fluid dynamics. We investigate the effects of release rate on overall plume dynamics such
as rise time and plume angle and compare CFD simulations with experimental observations. We find that
progress in sonar imaging now allows subsea gas plumes to be visualized in detail and may prove a very useful
tool in field situations. The CFD model captures all the main features observed in the experiments.
fluid dynamics (CFD) simulations. In the event of subsea loss of containment of either natural gas or carbon
dioxide it is important to understand how the gas disperses in the ocean column and acends to the surface where
it can pose a risk to installations, vessels and on-board personnel. Risks factors include; fire and explosion
hazards, sudden and persistant hydrodynamic loads, in case of carbon dioxide the hazard of asphyxiation, and
in case of a gas rich in hydrogen sulfide toxic effects. Thus, it is routine to conduct atmospheric dispersion
simulations to estimate size of exclusion zones on rigs and safe stand-off distances for vessels in case of loss of
containment. Atmospheric dispersion is today in some respects seen as a routine and mature exsersize. Howerver,
subsea dispersion of gas and surfacing of this gas is still an area with limited data and large uncertainties
with respect to a number of factors (Olsen & Skjetne(2016-A)); transient release rates, depth, mass transfer
to ocean, degassing at ocean surface. A key obstacle to obtaining good data has been the challenge of subsea
imaging and characterization of large transient bubble plumes structures. Here we utilize recent developments
in 3D temporal sonar imaging obtained for starting plumes from 30 meters depth with predictions obtained
using computational fluid dynamics. We investigate the effects of release rate on overall plume dynamics such
as rise time and plume angle and compare CFD simulations with experimental observations. We find that
progress in sonar imaging now allows subsea gas plumes to be visualized in detail and may prove a very useful
tool in field situations. The CFD model captures all the main features observed in the experiments.
