Feasibility of Using a Borehole Gravity Tool to Detect Fracture Systems at Reservoir Scale

Abstract In mature oil fields, water, steam, natural gas, and carbon dioxide are purposely injected into a reservoir to increase pore pressure on the oil, to lower the viscosity of heavy oil, and, ultimately, to increase the oil recovery factor. The placement of oil, water, and gas within the reservoir is highly dynamic and not easily predicted. Therefore, reservoir monitoring is key to improving sweep efficiency and oil recovery. Density is an important physical property for inferring oil, water, and gas saturation in a rock. Due to the density differences of oil, water, and gas, the saturation variation of each phase will have a direct impact on the measured gravity fields. The conventional well logging tool, gamma-gamma density, has a maximum depth of investigation (DOI) of 6 to 8in in open hole and greatly reduced accuracy in cased hole. A gravimeter, in contrast, yields a density measurement with inherently large DOI, where the DOI is determined by the vertical separation between pairs of differenced measurements of the acceleration due to gravity. For example, the DOI resulting from a pair of measurements separated by 3m will be approximately 4.5m. In this paper, we have conducted extensive simulations of using borehole gravity fields to monitor water flooding from injectors. In one scenario, we considered fracture systems within the reservoir typically identified by seismic methods; however, their impact on water flooding is not fully understood. Borehole imaging and production logging measurements clearly identify where fractures are along the borehole. However, due to their limited DOI, there is no clear picture about fractures farther from the borehole. Because the fractures are most probably filled with water and rock other than formation rock, which have a density different from that of the surrounding formation matrix, the borehole gravity field with its large DOI could provide useful information. To understand the response and the sensitivity of the borehole gravity measurements, we constructed a reservoir model with fractures of different sizes and locations within the modeled volume. From the lithology and fluid properties, density information was computed and propagated in the model. A vertical well was used as the survey location of the borehole gravity measurements. Both gravity and differential fields were calculated. The simulations were designed to address the following questions: Does the borehole gravity measurement have enough sensitivity to detect fractures within a reservoir?How far away from a borehole can fractures be accurately identified? This paper will present our simulation results and discuss their implication for reservoir fluid-front monitoring. Although borehole gravity measurements have not been used in reservoir monitoring, this study's results demonstrate the advantage of data acquisition with large DOI, which provides a huge operational benefit and cost-effective measurements to manage oil fields.