A Proven Squeeze-Cementing Technique in a Dolomite Reservoir

Abstract

Data on squeezed injection wells in a Grayburg-San Andres reservoir of West Texas indicate that a high final standing pressure is not a prerequisite for a successful squeeze. Profiles made a year after repair indicate that the wells have maintained the desired injectivity profile despite continued injection at fracturing pressure. Introduction The reservoir being discussed is located in West Texas and produces from the Grayburg-San Andres dolomites of Permian age at an average depth of 3,400 ft. The formation has low porosity and permeability (average 8 percent and less than 1 md, respectively) and is vertically fractured. Gross pay thickness ranges from 100 to 400 ft. Connote water saturations approximate 40 percent. The field, discovered in the 1930's, remained small until the advent of hydraulic fracturing in the 1950's. By 1960, development on 20-acre spacing was virtually complete. Completion Practices Except in the early wells, primary completion practices consisted of setting 5 1/2-in. casing at total depth and circulating cement. Float collars were used, and in many of the wells scratchers and centralizers were put over the pay section. The common cement put over the pay section. The common cement program consisted of API Class C-HSR 4-percent gel program consisted of API Class C-HSR 4-percent gel over the pay and Class C 8-percent gel as a filler to the surface. All of the wells were hydraulically fractured with lease or refined oil on completion. All of the jobs were one stage. Volumes ranged from 5,000 to 30,000 gal at rates from 15 to 30 bbl/min down the annulus between tubing and casing. Sand concentrations varied from 1 1/2 to 4 lb/gal. Acid was spotted in the hole before fracturing commenced. The majority of the wells were perforated either with one 30- to 60-ft interval at the base of the gross pay section or "blanket" perforated over the whole zone. Some wells have 4 shots/ft over intervals as great as 300 ft. Detection of Wellbore Channeling Pilot waterflooding operations in this reservoir were Pilot waterflooding operations in this reservoir were not successful. Other than two cases of direct communication to producing wells, no effect was observed. Several sets of radioactive tracer surveys were run, all of which showed that the water was confined to the desired intervals. No channeling was detected. Detailed studies were made to review primary performance and evaluate locations for one or more additional performance and evaluate locations for one or more additional pilots. In the course of these studies it became evident pilots. In the course of these studies it became evident from a review of the production history, workover success, bottom-hole pressures, and core analyses that the wells that were initially perforated low were in communication with other portions of the pay section. Failure of primary cement jobs or connection of vertical fractures in the formation were believed to be responsible for the communication. Based on the inference of wide-scale mechanical problems from these data, temperature logs were then run on the pilot water injection wells. These surveys showed that most of the injected water was being lost into nonproductive intervals in all of the injection wells. The search was then begun for a suitable squeeze-cementing technique to confine the water to the pay zone at the wellbore in hopes that this would provide the key to successful waterflooding. Subsequent performance data have shown conclusively that the problem is at or near the injection wellbore. JPT P. 1341