Multiwell Experiment Research Articles

Determination of Widths of Meander-Belt Sandstone Reservoirs from Vertical Downhole Data: ABSTRACT

Once it has been determined that a meandering fluvial model is applicable to a formation, paleohydrologic reconstructions can be applied to downhole measurements to derive sand body widths. The nonmarine part of the Mesaverde Formation in the east-central part of the Piceance Creek basin of northwestern Colorado was deposited in a predominantly meandering fluvial End_Page 505------------------------------ system. This interpretation is based on the presence of lateral accretion beds and multi-story sand bodies in outcrops along the Grand Hogback. It is supported by fining-upward trends in cores and gamma ray logs from the wells of the U.S. Department of Energy's Multi-Well Experiment being conducted near Rifle, Colorado. Paleochannel depths, recognized from the heights of fining-upward trends in cores, can be converted to channel widths using Leeder's 1973 formula. The resulting channel widths are used to calculate meander belt amplitudes (sandstone body widths) from relationships derived from Leopold and Wolman's 1960 empirical data. These numbers can be compared with sandstone body widths derived by two other methods. (1) Point-bar dimensions measured in outcrops can be used to calculate channel widths, which are then converted to meander-belt amplitudes as described above. (2) The MWX-2 well is offset from MWX-1 by 135 ft (41 m) at the surface, allowing for positive well-to-well correlation of sandstones. The percentage of sandstones which are penetrated by both wells is used as a probability to derive an average sand body width. These last two methods give compatible results. The first method described, however, predicts comparatively narrow widths, suggesting that channel depths derived from fining-upward sequences in cores are not fully representative of ancient channel depths due to incomplete preservation. A preservation potential factor may be added to Leeder's formula for working with ancient sediments, so that the formula gives comparable results to the other two methods. End_of_Article - Last_Page 506------------

Vitrinite Reflectance and Temperature Gradient Models Applied at a Site in Piceance Basin, Colorado: ABSTRACT

Two adjacent DOE multiwell experiment (MWX) boreholes were drilled to a depth of 2,520 m (8,268 ft) below the valley of the Colorado River near Rifle, Colorado. High quality vitrinite reflectance data ranging from 0.88 to 2.07% were obtained from cores from the Upper Cretaceous coaly interval (1,340 to 2,423 m; 4,396 to 7,949 ft). Remnants of basalt flows that cap highlands above the valley are 9 m.y. old, and 1,500 m (4,921 ft) of Eocene and Pliocene sediments have apparently been eroded from above the site by the Colorado River since the basalt was extruded. The excellent reflectance data and the geologic setting allow a test of three ways to evaluate maturation of vitrinite. Figure End_Page 427------------------------------ Logarithms of the reflectance values form a linear trend with depth. Extrapolation of this trend to 0.20% Ro, a value thought by some to mark the normal surface intercept, reaches only 460 m (1,509 ft) above present ground, instead of the 1,500 m (4,921 ft) predicted on the basis of postulated erosion. Such a deep intercept would indicate that maximum temperatures were attained much more recently than the volcanism (and subsequent to removal of 1,040 m [3,412 ft] of sediment). If the plot is extended up to the +1,500 m (+4,921 ft) pre-erosion level, the intercept is only 0.08% Ro, leading to an unlikely paleotemperature situation. Time-independent models of the general dependence of reflectance on temperature yield a paleotemperature gradient of about 115°C/km and surface intercepts (20°C) of -350 to +250 m (-1,148 to +820 ft) relative to the present ground level. These models require that a strong heating occurred at the site at nearly the present time, and are not in accord with available facts. Time-dependent models give a paleotemperature gradient of 50° to 65°C/km in the sampled interval (41 to 54°C/km in the overlying non-coaly section). Intercepts of 20°C are +1,300 to +1,700 m (+4,265 to +5,577 ft) above the present ground. These models agree logically with the present 46°C/km gradient across the sampled interval and the removal of 1,500 m (4,921 ft) of overburden, with maximum temperature just a bit after maximum burial. Of the three approaches tried, the time-dependent model is the only one which works in this situation. End_of_Article - Last_Page 428------------