S U M M A R YObserved deformation is sensitive to a changing stress field within the Earth. However, there are several impediments to a direct inversion of geodetic measurements for changes in stress. Estimating six independent components of stress change from a smaller number of displacement or strain components is inherently non-unique. The reliance upon surface measurements leads to a loss of resolution, due to the attenuation of higher spatial frequencies in the displacement field with distance from a source. We adopt a technique suited to the estimation of stress changes due to the injection and/or withdrawal of fluids at depth. In this approach the surface displacement data provides an estimate of the volume change responsible for the deformation, rather than stress changes themselves. The inversion for volume change is constrained by the fluid fluxes into and out of the reservoir. The distribution of volume change is used to calculate the displacements in the region above the reservoir. Estimates of stress change follow from differentiating the displacement field in conjunction with a geomechanical model of the overburden. We apply the technique to Interferometric Synthetic Aperture Radar (InSAR) observations gathered over a petroleum reservoir in the San Joaquin Valley of California. An analysis of the InSAR range changes reveals that the stress field in the overburden varies rapidly both in space and in time. The inferred stress variations are found to be compatible with the documented failure of a well in the field.

The ability to determine oil saturation changes through time in the Kern River field is critical to heat management and optimization of reserves depletion. The Kern River reservoir is a complex multi-layered fresh water braided river deposit undergoing varying stages of thermal recovery. In the past, the practice had been to selectively core when new well locations were identified. An alternative to open hole core saturation measurements is the Carbon Oxygen Ratio (COR) log, which yields accurate and repeatable data that can be used to identify and monitor reserves depletion. In this field, time-lapse 4D oil saturation images are being generated from a grid of several hundred observation wells. Over 360,000 feet of COR data have been acquired. The project is realizing accuracies comparable to core data — bore out from a database containing 36,000 feet of core. In 1993, nine observation wells were drilled and cored. After casing was cemented in place, the wells were logged immediately to reduce uncertainties associated with subsequent production. Full field monitor logging was implemented in 1996 and again in 1997. Small changes in the COR oil saturation measurement over time are being monitored with a measurement precision of +/- 3.2 saturation units (su) with a confidence level of 95%. The highly accurate and precise data were made possible in part by reservoir conditions, well bore conditions, and a process improvement program incorporating all major steps from acquisition to data processing. A shared database tracked progress and monitored quality throughout the project. The project deliverables were in digital form only. Log-to-core data comparisons, steps to determine precision and accuracy, as well as time-lapse 4D images showing reservoir dynamics, will be discussed.

We studied crosswell electromagnetic and seismic images of three oil‐saturated intervals within a southern California heavy oil field undergoing steam flood. The crosswell survey is located in a portion of the field where one well is in a “cold spot,” resulting in differing steam propagation within the three units. Log analysis shows linear or second‐order polynomial relationships (with correlation coefficients greater than 0.7) between electrical conductivity and water saturation, porosity, and clay content; whereas only a weakly linear relationship can be found between velocity and temperature in the lower unit studied. Crosswell seismic data is used to produce a velocity tomogram and a reflection section, and crosswell electromagnetic data is used to produce a conductivity section and derived porosity and water saturation. The seismic velocities from the tomograms show lateral variations consistent with the lateral variations in temperature seen in observation wells on either side of the crosswell section. The continuity and disruption of seismic reflections coincide with zones of continuous and variable porosity and water saturation as produced from the crosswell electromagnetic inverted conductivity section, and the derived regression fits between conductivity and porosity/water saturation. Seismic velocities, reflections, electrical conductivity, and the derived porosity and water saturation sections all contribute to explaining the observed lateral temperature variations between the wells within the three reservoir units. The unit with high steam content has low water saturation, high porosity, and laterally continuous low‐velocity and seismic reflections, consistent with no barriers to flow. The upper unit, where steam breaks through later in the experiment, also has high porosity and laterally continuous seismic reflections. However, it shows a velocity gradient between the hot and cold wells consistent with the lack of steam in the cold well at the time of the experiment. The middle unit, in which steam never reaches the cold well during the experiment, has the highest water saturation and the largest zone of reduced porosity and disrupted seismic reflections, indicating a possible barrierto flow.

Ultra-lightweight cements have been used throughout the industry for many years as a specialty cement for application in areas of weak formations. Typical lightweight slurries use hollow ceramic microspheres or nitrogen and other low specific gravity materials to maintain light slurry densities while providing good compressive strength development. The one problem with these systems has been the evaluation of the job via sonic tools. This paper will introduce a new technique to allow an accurate interpretation of a sonic cement evaluation and will be demonstrated from laboratory model studies. Light cement evaluation studies show a statistical variance technique to be the preferred method of evaluation in its class. Standard methods of cement evaluation rely on the contrast in acoustic properties of materials behind pipe. The Cement Bond Log (CBL) acoustic amplitude is used to derive compressive strength. The ultrasonic tool uses acoustic impedance to derive compressive strength. However, in the process of lightening the cement slurry, the acoustic properties of the cement can approach the acoustic properties of a fluid. Modeled values were compared to measured values of various slurries from 16.4 lb/gal to 9.0 lb/gal, confirming the trend from high acoustic impedance contrast of a cement to a drilling fluid in the case of a 16.4 lb/gal slurry to a negative contrast in the case of a 9.0 lb/gal slurry. Situations can arrive where drilling mud will be displaced by a cement slurry of the same density. Standard cement evaluation techniques would not be able to identify the cement in this case because of a lack of acoustic impedance contrast. Another technique is available for low-density cement evaluation that identifies the cement from the fluid behind pipe by quantifying the other statistical variance of the acoustic impedance measurement. It will be demonstrated that there is a difference in the impedance measurement statistically between a solid crystalline cement and fluid, even though the absolute values are the same. Additional work will demonstrate the difference between designed lightweight cement with nitrogen and cement that has been subjected to gas migration.

This paper presents the problems of evaluating the integrity of gas-channeled cement sheaths and with ultrasonic and sonic bond tools utilized extensively in field operations. Findings are based on full-scale models intentionally gaschanneled, logged by standard ultrasonic and sonic bond tools, then exposed for visual inspection. Possible methods to overcome the discovered problems of cement evaluation with bond tools are presented so that alternate methods of determining cement integrity can be used.Findings presented show that normal cement bond logs will not always detect gas channeling in the cement sheath unless they are created by large gas influx rates which increases the probability for the channels to contact the casing. Laboratory findings verify field examples of logs showing gas-influx cement, but physical inspection indicates distinct gas channels rather than small gas bubble entrainment. The variance method is utilized with averaging impedance values to distinguish between gas channeling and foam cements. Cement jobs in the industry are typicallyReferences and illustrations at end of paper. evaluated with sonic and/or ultrasonic tools to indicate hydraulic isolation. Gas channels in the cement may compromise hydraulic seal if the channel communicates between uneconomical zones and production perforations. This paper provides recommendations to allow evaluation of gas channel cement for a squeezelno squeeze decision from log responses and physical inspection of the models.