--> Time-Lapse Imaging of Heavy Oil Reservoirs at Shallow and Deep Using Ultra-Stable Seismic Sources

AAPG Annual Convention and Exhibition

Datapages, Inc.Print this page

Time-Lapse Imaging of Heavy Oil Reservoirs at Shallow and Deep Using Ultra-Stable Seismic Sources

Abstract

For EOR, we have used an innovative technology called seismic ACROSS (Accurately Controlled and Routinely Operated Signal System). The ACROSS seismic source gives eccentric force using the rotation of a mass adjusted to the GPS time base. The frequency sweep is 10 Hz to 50 Hz generating force is 40f tons at 50Hz. We can simultaneously generate vertical and horizontal vibrations by calculation. We assume a few ACROSS seismic source(s) and 2D or 3D geophone array. Using residual waveforms of before and after the injection of vapor or super critical CO2 to reservoirs, we can image the time-lapse of reservoir softening by the “reciprocity principle of Green function” similar to reverse time method. In our previous studies, we carried out simulations at heavy oil, shale gas and CCS and three field studied in Japan and Saudi Arabia. The real applications of our method to the oil fields might require inexpensive installation costs and similar quality of reservoir imaging as an ordinary 4D seismic survey in which 25m source and receiver spacing are commonly used in contrast to a few sources and larger source and receiver spacing in our method. Coarse geophone spacing could greatly contribute to reduce the installation costs if we can prove to get comparable resolution to 3D seismic survey. To prove the usefulness of our method applying to the real oil/gas exploration, we carried out new simulations of reservoirs at 2km and 200m depths. In the 2 km depth model, we used only one ACROSS seismic source and a geophone array of 200m spacing in 2D and 3D grids. The result of 4D simulation gave a precise retrieval image even though we used only one seismic source and an array of 200m geophone spacing. When we compare 5m and 200m geophone spacing in 2D model, we cannot identify significant difference between two spacing models. The second case is very shallow heavy oil reservoir. We tested 100m geophone grids and one or two ACROSS source(s) to image the small size of target reservoir (l=20m, w=20m, d=10 m) at 200m depth. Even the two sources are 850m distance from the target reservoir, a tiny target is well imaged. Because the image obtained by 106 and 25 receivers do not show any significant changes, retrieved image are mostly determined by receives just above the target. In conclusion, our time-lapse approach enables us to drastically reduce the costs for the design of monitoring system maintaining good imaging quality for temporal change of reservoirs.