The borehole gravity meter (BHGM) is recognized as an important logging tool for obtaining formation bulk density. In general, however, the difference between two gravity observations vertically separated in a well leads to an apparent and not the actual bulk density. BHGM‐derived apparent densities are equal to the formation bulk densities when the instrument passes through beds which are horizontal, infinitely extended laterally, uniformly thick, and constant in density. For many applications, departures from these assumed conditions are so slight that their effects can be ignored, and the BHGM essentially yields bulk density with a large radius of investigation. In the presence of anomalous masses, significant distortion in formation bulk density is possible. The apparent density anomaly produced in the well by an elongated, offset density contrast is proportional to the angle subtended by the density‐change interface. For a density‐change boundary having circular symmetry with respect to the well, the apparent density anomaly at the center of the bed is proportional to the sine of the subtended angle. Because the distortion in bulk density is the same above a horizontal boundary as it is just below (in the limit, at the boundary, for a normally incident well), an abrupt change in apparent density is equal to the real density change at the boundary. This change in density, termed “the Poisson jump,” is independent of geometry; our ability to measure it, however, is a function of station location with respect to the geologic bodies. Two methods are suggested for obtaining bulk densities from BHGM apparent densities: (1) by obtaining two stations just outside as well as just within the zone of interest, the Poisson jump can be approximated and added to an independent density source (e.g., the gamma‐gamma log), and (2) the apparent density anomaly within the formation of interest can be derived by modeling (perhaps based on seismic or well data) and added to the BHGM‐determined densities. Thinner beds can be studied with the BHGM than generally believed, even with much greater station spacing.
Dynamic Planning of Bicycle Stations in Dockless Public Bicycle-sharing System Using Gated Graph Neural Network
