Larry Mewhort

Delineation of Steam Flood Using Seismic Attenuation

The possible use of attenuation measurements for time-lapse seismic monitoring of an EOR steam flood project in Saskatchewan, Canada is investigated. A VSP survey was used to calculate Q. These values were input to a synthetic seismogram attenuation modeling program that showed there should be an observable increase in attenuation after steam injection. Two seismic lines shot at the same location nine years apart were analyzed to see if attenuation anomalies were apparent. The results indicate a strong anomaly on the recent seismic line that is consistent with the location of steam injection. A weaker anomaly on the older line is consistent with the amount of steam injected at that time. Theoretical and laboratory analyses of compressional velocity as a function of changes in temperature, pressure, fluid type and fluid phase suggest there should be a measurable effect on compressional wave amplitude, isochron, and frequency response.

The use and abuse of geostatistics

Geostatistical reservoir characterization techniques promise to improve the economics of oil and gas production through the integration of seismic, petrophysical and geologic techniques. Unfortunately, these methods have become veiled with mysterious jargon and the use of geostatistics is too often restricted to a few “experts.” In reality, geostatistical tools are quite simple to use and can be easily mastered. However, these new techniques can only supplement and not replace the correct practice of geophysics, geology, and reservoir engineering.

Gabor Deconvolution: Real And Synthetic Data Experiences

As the search for subtle hydrocarbon traps intensifies, the seismic processing world is driven to extract increasing bandwidth from the raw data. To this end, various high resolution processing schemes have been proposed, all of which seek to restore those high frequency Fourier components of the signal which have been attenuated via earth filtering processes such as anelastic attenuation and stratigraphic filtering. Among these approaches, that one consisting of a single pass of surface-consistent, minimum phase deconvolution followed by multiple passes of prestack and/or poststack time-variant spectral whitening (TVSW) has become quite popular. Although years of experience have shown that the TVSW algorithm is amazingly robust, detractors are quick to point out that it lacks a certain amount of theoretical rigour. This criticism stems from the fact that non-stationary earth filtering processes (whose residual effects TVSW attempts to remove) are minimum phase, while the algorithm effectively constructs temporally localized zero-phase operators and therefore produces results which may contain residual phase errors. By contrast, Gabor deconvolution (Margrave et al., 2002, Margrave et al., 2003, Margrave et al., 2005) constructs temporally localized minimum phase, rather than zero phase, operators, and as such it offers a more theoretically rigorous alternative to this conventional high resolution flow.

A Modeling Study for Imaging in Structurally Complex Media: Case History

Two dimensional finite difference seismic modeling of a complex geologic structure was used to aid in seismic interpretation and the optimization of future acquisition and processing parameters. The depth/velocity model was derived from a balanced geological section in North Eastern British Columbia foothill inferred from a 1998 2D-seismic line, acquired by Husky Energy Corporation, and 5 adjacent wells drilled from 1981 to 2003. The model contains a velocity inversion, rough topography and characteristic structures found in NE BC.

Seismic characterization of thermal recovery processes at Pikes Peak

Seismic moni tor ing of thermal recovery processes h a s b e c o m e a n e s t a b l i s h e d technology over recent years. This technique provides information related to the aerial distribution of injected steam in the reservoir and this information can be used to assist in drilling infill wells that contact unswept areas t o i m p r o v e u l t i m a t e r e c o v e r y . Seismic monitoring is usually performed in a timelapse fashion. An initial survey is recorded prior to steam injection to measure spatial variations of the seismic signal caused by geologic variations. Subsequent surveys are recorded after the recovery process has been ac t ive for a per iod of t ime . Di f fe rences between the initial survey and the later survey a r e u s e d t o d e t e r m i n e t h e a r e a o f t h e reservoir that has been affected by thermal recovery. However, recent research (Eastwood et al, 1996) has shown that seismic monitoring can be performed on the basis of a single seismic survey, provided the geologic variations and the production-induced seismic anomalies can be statistically separated using a combination of seismic attributes. We have success fu l ly app l ied th i s t echn ique to a t h e r m a l r e c o v e r y p r o j e c t f r o m W e s t e r n Canada to map the probable distribution of injected steam in a reservoir that had been under active steam-drive for more than 10 years before the seismic survey was recorded. These results have been validated through a comparison with production data.