Challenges in reservoir forecasting

The combination of geostatistics-based numerical geological models and finite difference flow simulation has improved our ability to predict reservoir performance. The main contribution of geostatistical modeling has been more realistic representations of reservoir heterogeneity. Our understanding of the physics of fluid flow in porous media is reasonably captured by flow simulators in common usage. Notwithstanding the increasing application and success of geostatistics and flow simulation there remain many important challenges in reservoir forecasting. This application has alerted geoscientists and physicists that geostatistical/flow models in many respects, are, engineering approximations to thereal spatial distribution andreal flow processes. This paper reviews current research directions and presents some new ideas of where reserach could be focused to improve our ability to model geological features, model flow processes, and, ultimately, improve reservoir performance predictions.     

Theory for the semi-analytical calculation of oil recovery and effective relative permeabilities using streamtubes

A semi-analytical method has been developed for calculating oil recovery in two and three dimensions, and for calculating effective relative permeabilities for coarse grids. The calculations are based on the assumption that the effects of a changing mobility field can be accounted for by using fixed streamtube geometries with flowrates updated to account for the changing mobility distribution. The single-phase pressure distribution from a numerical solution of Laplace's equation is used to calculate the pressure distribution for a two-phase flow based on a mapping of the solution of the Buckley-Leverett equation onto the streamtubes derived from the single-phase solution. The displacement calculations for oil recovery are based on theory previously developed by Dykstra and Parsons, extended to include the effects of spatially varying permeability and continuously changing mobilities, as occurs in solutions of the Buckley-Leverett equation for typical values of the mobility ratio. This idea has also been extended to the calculation of effective relative permeabilities for coarse-grid simulation and finally establishes the proper rules for averaging the results of fine-grid numerical simulations of two-phase flow for the definition of effective two-phase flow properties on coarse grids. These calculations have been generalized to three-dimensional flows by the simple device of conceptually inserting a gridded plane across the flow and defining each streamtube at that location as those streamlines which pass through any one of the grid cells. When combined with time-of-flight calculations from the gridded plane to both the producer and injector, the distribution of pore volume along each streamtube can be calculated. This information, combined with a tabulation of the single-phase, steady-state pressure distribution along each streamtube, provides all of the information needed for the semi-analytical calculation of oil recovery and effective flow properties in three-dimensional flows. © 1997 Elsevier Science Ltd. All rights reserved     

Dynamic Three-Dimensional Tomography of the Solar Corona

Empirical, three-dimensional electron-density maps of the solar corona can be tomographically reconstructed using polarized-brightness images measured from ground- and space-based observatories. Current methods for computing these reconstructions require the assumption that the structure of the corona is unchanging with time. We present the first global reconstructions that do away with this static assumption and, as a result, allow for a more accurate empirical determination of the dynamic solar corona. We compare the new dynamic reconstructions of the coronal density during February 2008 to a sequence of static reconstructions. We find that the new dynamic reconstructions are less prone to certain computational artifacts that may plague the static reconstructions. In addition, these benefits come without a significant increase in computational cost.     

The UKIRT wide field camera ZYJHK photometric system: calibration from 2MASS

In this paper, we describe the photometric calibration of data taken with the near-infrared Wide Field Camera (WFCAM) on the United Kingdom Infrared Telescope (UKIRT). The broad-band ZYJHK data are directly calibrated from Two-Micron All-Sky Survey (2MASS) point sources which are abundant in every WFCAM pointing. We perform an analysis of spatial systematics in the photometric calibration, both inter- and intradetector show that these are present at up to the ∼5 per cent level in WFCAM. Although the causes of these systematics are not yet fully understood, a method for their removal is developed and tested. Following the application of the correction procedure, the photometric calibration of WFCAM is found to be accurate to ≃1.5 per cent for the JHK bands and 2 per cent for the ZY bands, meeting the survey requirements. We investigate the transformations between the 2MASS and WFCAM systems and find that the Z and Y calibrations are sensitive to the effects of interstellar reddening for large values of   E ( B − V )'  , but that the JHK filters remain largely unaffected. We measure a small correction to the WFCAM Y -band photometry required to place WFCAM on a Vega system, and investigate WFCAM measurements of published standard stars from the list of UKIRT faint standards. Finally, we present empirically determined throughput measurements for WFCAM.     

Kinematics of SDSS subdwarfs: structure and substructure of the Milky Way halo

We construct a new sample of ∼1700 solar neighbourhood halo subdwarfs from the Sloan Digital Sky Survey (SDSS), selected using a reduced proper-motion diagram. Radial velocities come from the SDSS spectra and proper motions from the light-motion curve catalogue of Bramich et al. Using a photometric parallax relation to estimate distances gives us the full phase-space coordinates. Typical velocity errors are in the range  30–50 km s⁻¹  . This halo sample is one of the largest constructed to date and the disc contamination is at a level of ≲1 per cent. This enables us to calculate the halo velocity dispersion to excellent accuracy. We find that the velocity dispersion tensor is aligned in spherical polar coordinates and that  (σ _r , σ_φ, σ_θ) = (143 ± 2, 82 ± 2, 77 ± 2) km s⁻¹  . The stellar halo exhibits no net rotation, although the distribution of   v _φ  shows tentative evidence for asymmetry. The kinematics are consistent with a mildly flattened stellar density falling with distance like   r ^−3.75  .
Using the full phase-space coordinates, we look for signs of kinematic substructure in the stellar halo. We find evidence for four discrete overdensities localized in angular momentum and suggest that they may be possible accretion remnants. The most prominent is the solar neighbourhood stream previously identified by Helmi et al., but the remaining three are new. One of these overdensities is potentially associated with a group of four globular clusters (NGC 5466, NGC 6934, M2 and M13) and raises the possibility that these could have been accreted as part of a much larger progenitor.     

Discs in early-type lensing galaxies: effects on magnification ratios and measurements of H0

Observations of early-type galaxies, both in the local Universe and in clusters at medium redshifts, suggest that these galaxies often contain discs or disc-like structures. Using the results of Kelson et al. for the incidence of disc-components among the galaxies in the redshift   z = 0.33  cluster CL 1358+62, we investigate the effect of disc structures on the lensing properties of early-type galaxies. Statistical properties, like magnification cross-sections and the expected number of quad (four-image) lens systems, are not affected greatly by the inclusion of discs that contain less than ∼10 per cent of the total stellar mass. However, the properties of individual lens systems are affected. We estimate that 10–30 per cent of all quad-lens systems, with early-type deflector galaxies, would be affected measurably by the presence of disc components. Intriguingly, the image magnification ratios are altered significantly. The amplitude of the predicted change is sufficient to explain the observed magnification ratios in systems like B1422+231 without requiring compact substructure. Furthermore, time-delays between images also change; fitting a bulge-only model to early-type lenses that in fact contain a disc would yield a value of the Hubble constant H ₀ that is systematically too low by ∼25 per cent.