Computer Processing of Log Data Improves Production In Chaveroo Field

Abstract

Computer log analysis has proved quite successful in the Chaveroo field of Roosevelt County, N.M., in a carbonate section of low permeabilities, low porosities and variable water saturations. Production in this field occurs at about 4,300 ft in the San Andres formation, a Permian carbonate with a mixture of dolomite, anhydrite and gypsum, and lesser amounts of silica, limestone and silt. Sonic, density and neutron logs, all of which respond differently to porosity and lithology, enable a mathematical solution that defines the fractions of major lithologic components and gives the correct porosity. The Laterolog and Microlaterolog are used to solve for a movable oil relationship. The computer output is a log of bulk volume percentages of porosity, dolomite, anhydrite, gypsum and silica, and a movable oil plot. The record shows that wells completed on the basis of this computer log analysis statistically have higher initial productions and lower water cuts than wells completed on the basis of less diagnostic logging techniques. Introduction The Chaveroo pool is characterized by conditions generally considered hostile to effective log interpretation. Porosities are low-about 5 percent in most producing intervals. The intervals are predominantly dolomite, but contain unpredictably varying fractions of anhydrite, gypsum, chert, silt and limestone. Formation resistivities are high, and most wells are drilled with salt mud. However, in spite of these difficult conditions, modern logs and interpretation methods are giving information needed for efficient completion in the Chaveroo pool. The more efficient completions are obtained by avoiding water-productive intervals interspersed among the pay section of the pool. The produced water fraction is important in Chaveroo because it usually determines whether a well will flow or will require pumping. Thus, better completions are possible when water-productive intervals are distinguished from pay zones. Logs both define pay sections and identify water-productive intervals. Porosity information from logs is used with resistivity data to define fluid saturations. However, because the lithology is variable, no single log defines porosity with sufficient accuracy to reliably distinguish between pay and water. Sonic, neutron and density logs, all of which respond differently to porosity and lithology, are used in a mathematical solution that identifies major lithologic components and thus affords a reliable value of porosity. Office-based electronic computers are used to perform the mathematical solutions. The output of the computer is a log of percentages of major lithologic components and porosity, and an index of movable oil and water saturation. Statistics show that wells selectively completed on the basis of these computer logs are more efficient producers than other wells of the Chaveroo pool. These wells generally flow with higher initial oil production and produce less water. The logs and interpretation methods presented in this article are not entirely new. Earlier papers have described the primary logs, the basic interpretation methods and typical applications of computers to log studies. This study is presented to show how the various logs and methods are combined for improved well completion in the Chaveroo pool. The methods are not restricted to Chaveroo; they are similarly appropriate for other areas of complex lithology. Chaveroo Field Chaveroo field straddles the line between Chaves and Roosevelt Counties of eastern New Mexico (Fig. 1). Production is from the Slaughter zone of the San Andres formation. The main producing interval is about 650 ft below the top of the San Andres, and is at an approximate depth of 4,150 ft. The Slaughter zone, situated between two massive anhydrite sections, is 200 to 250 ft thick. It is described as a fractured, vuggy dolomite having some intercrystalline and oolitic porosity. In addition to interbedded and nodular anhydrite, the zone also contains variable but lesseramounts of gypsum, chert, silt and limestone. JPT P. 889ˆ

Data Report: Formation MicroScanner Imagery of Lower Cretaceous and Jurassic Sediments from the Western Pacific (Site 801)

“…The basic theory for the modeling of compositional components from logging data was developed by Savre (1963), and further improved by Burke et al (1967) and by Dawnson-Grove and Palmer (1968). In this model, the log response for any zone is related to the sum of the proportions of the mineralogical/porosity components, each multiplied by the appropriate response coefficients of the different logging instruments.…”

Interpretation of Site 948 logging-while-drilling data: mineralogical inversion of log data constrained by sample analyses

“…Each log provides a single equation. This matrix formulation is a linear model introduced by Savre (1963) and further developed by Burke et al (1967). More extensive details on this methodology are given by Doveton (1994).…”

Openhole Crossplot Concepts A Powerful Technique in Well Log Analysis

“…Such crossplots (Fig. 1) provide shaliness definition, evaporite determination (anhydrite, halite, and polyhalite; gypsum in clean, nongas-bearing zones), poor porosity resolution, and definition of Psh, ilt sh' and M sh ' Light hydrocarbons, gas, and gypsum shift the points toward the northwest; shales and compaction effects shift the points toward the east; secondary porosity shifts the trend toward the west; and borehole rugosity and washouts shift the trend toward the north.…”