Atmospheric monitoring and verification technologies for CO2 geosequestration

Abstract The paper describes various techniques for measuring emissions to the atmosphere from geologically stored carbon dioxide, from point, line and area sources at scales of metres to several kilometres. Flux chambers are suitable for measuring small leakage rates from sources at known locations but many samples are required because of large spatial heterogeneity in the fluxes. Micrometeorological eddy covariance, relaxed eddy accumulation and flux-gradient techniques are suitable for measuring leakage from large area sources, while integrated horizontal mass balance, tracer methods and plume dispersion approaches are applicable for line and point sources. Distinguishing between leakage signals and natural fluctuations in CO 2 concentrations due to biogenic sources pose significant challenges and the use of naturally occurring tracers such as CO 2 isotopologues or introduced tracers such as SF 6 added to the sequestered CO 2 will assist with this problem. Forward Lagrangian dispersion calculations showed that CO 2 concentrations 0–80 m downwind of a point source would be readily detectable above all natural variations for point sources >0.3 g CO 2  s −1 (about 10 tonnes of CO 2 per year). The inverse problem involves solving for the unknown emission rate from measured wind fields and down wind concentration perturbations. An optimum monitoring strategy for inverse analysis will require continuous measurements of CO 2 and tracer compounds upwind and downwind of the possible leak location, coupled with transport modelling to determine leakage fluxes, and to differentiate them from other sources. Computations using The air pollution model (TAPM) showed that expected perturbations in CO 2 concentrations at distances of several hundred metres from a leak of 32 g CO 2  s −1 (about 1000 tonnes CO 2 per year, or about 0.01% per year of a typical amount to be stored) will be detectable, but this anomaly will be very small compared to natural variations, thereby complicating the inverse analysis. While the techniques canvassed here have proven successful for measuring fluxes in other applications, none has yet been demonstrated for geosequestration. The next step is to test them in the field.