UGS Time Lapse Gravity

AAPG ACE 2021.
Underground gas storages (UGS) are an important part of the gas infrastructure that ensures a stable and effective supply of yearly changing demand in natural gas. Effective UGS’s exploitation requires a good understanding of the reservoir geological model to maintain the technical exploitation regime. Time-lapse UGS monitoring is usually performed in inspection wells and via 4D seismic survey and incorporates 3D models of porosity and permeability, pressure, and volume of gas by dynamic modeling. But often 4D seismic is not available and well data are too sparse to build an adequate reservoir model. That was the case for the Dashava UGS, located in the Carpathian region of Ukraine. The UGS was created in the old gas field in 1987. 27 injection/withdrawal cycles resulted here one-third increase of cushion gas volume in the reservoir. 3D seismic was performed over approximately 2/3 of the UGS area, leaving a big part of it unexplored. This, together with a sparse well net in the unexplored area made dynamic modeling inapplicable. To identify the location of an excess cushion gas volume time-lapse high accuracy gravity measurements were performed. Four repeated gravity surveys (two series over the “empty” UGS and two – over the full UGS) were performed in 2012-2013 and resulted in a standard deviation of 4.3 mkGal. Results of repeated high-accuracy gravity measurements confirmed the presence of gravitational anomalies in the range of 12.0–24.0 mkGal, related to the change of the gas volume in the UGS. The physical base for interpretation is a dependence between an in situ gas density and reservoir pressure/gas volume in the reservoir. For the case illustrated we observed direct dependence between gravity anomaly and gas volume for 60% of gravity stations. For another 40% of stations, dependence was inverse. 4D joint inversion of gravity data with 3D seismic data and well data were used to create a 4D density model, models of gas pressure, and the gas volume in place. Models’ resolution was 1 m vertically and 50x50 m laterally. As a result of the inversion, we mapped the location of dynamic reservoirs in the peripheral part of the UGS, where cushion gas is accumulated. 4D models allowed to calculate gas volume within peripheral dynamic reservoirs, and to map working gas migration pathways from the storage wells to dynamic reservoirs. The relative misfit of gas pressure prediction in the 4D model was 3.7 % in the case of “empty” UGS and 3.5 % in the case of full UGS. The relative misfit of working gas volume prediction and official gas balance was 1.0 %. The study resulted in the identification of the dynamic reservoirs and pathways of cushion gas movement, that can be recommended for displacement by the gas with high density, for example, nitrogen, to extract part of the cushion gas volume and to block migration to the dynamic reservoirs in peripheral parts of the UGS.