No abstract

No abstract

Thirty seafloor gravity stations have been placed above the carbon dioxide [Formula: see text] injection site and producing gas reservoir at the Sleipner Øst Ty field. Gravity and depth measurements from 2002 and 2005 reveal vertical changes of the permanently deployed benchmarks, probably caused by seafloor erosion and biologic activity (fish). The original gravity data have been reprocessed, resulting in slightly different gravity-change values compared with earlier published results. Observed gravity changes are caused by height variances, gas production and water influx in the Ty Formation, and [Formula: see text] injection in the Utsira Formation. Simultaneous matches to models for these effects have been made. The latest simulation model of the Ty Formation was fitted by permitting a scale factor, and the gravity contribution from the [Formula: see text] plume was determined by using the plume geometry as observed in 4D seismic data and varying the average density. The best-fit vertical gravity gradient is [Formula: see text], and the response from the Ty Formation suggests more water influx than expected in the presurvey simulation model. The best-fit average density of [Formula: see text] is [Formula: see text]. Estimates of the reservoir temperature combined with the equation of state for [Formula: see text] indicate an upper bound on [Formula: see text] density of [Formula: see text]. The gravity data suggest a lower bound of [Formula: see text] at 95% confidence.

Changes with gravity over time have proven to be valuable for inferring subsurface density changes associated with production from oil and natural gas reservoirs. Such inferences allow the monitoring of moving fluid fronts in a reservoir and provide an opportunity to optimize production over the life of the reservoir. Our group began making time-lapse seafloor gravity and pressure measurements in 1998. To date, we have surveyed six fields offshore Norway; we have made three repeat surveys at one field and one repeat survey at another. We incorporated a land-gravity sensor into a remotely operated seafloor housing. Three such relative gravity sensors mounted in a single frame are carried by a remotely operated vehicle (ROV) to concrete benchmarks permanently placed on the seafloor. Reference benchmarks sited outside the reservoir boundaries are assumed to provide stable fiducial points. Typical surveys last from a few days to a few weeks and cover from 8 to 80 benchmarks, with multiple observations of each. In our earliest surveys, we obtained an intrasurvey repeatability of approximately [Formula: see text], but recently we have been achieving [Formula: see text] repeatability in gravity and approximately [Formula: see text] in benchmark depth (deduced from simultaneously recorded ambient seawater pressure). We attribute the improved precision to several operational factors, including the use of multiple gravity sensors, frequent benchmark reoccupation, precise relocation and orientation of the sensors, repeated calibrations on land, and minimization of vibrational and thermal perturbations to the sensors. We believe that high-precision time-lapse gravity monitoring can be used to track changes in the height of a gas-water contact in a flooded reservoir, with a precision of a few meters.