Cement Bond Logging Techniques - How They Compare and Some Variables Affecting Interpretation

Abstract

BADE, J.F., JUNIOR MEMBER AIME, SHELL OIL CO., BILLINGS, MONT. Abstract Approximately 250 cement bond logs have been run in the Cedar Creek anticline, Mont., since late 1960. Continuing investigation has produced several significant tool design changes. These are discussed, and the relative merits and shortcomings r each design noted. Variables affecting the amplitude of measured signals include gating arrangement, tool sensitivity, centering, cement density, thickness of cement sheath, casing size, logging fluid and formation velocities. Field examples illustrating these factors are presented along with suggested methods of recognizing and coping with them. As a further aid in evaluating the nature of the signals measured by the commercial devices, a log was run in conjunction with four commercially available logs which photographed the received acoustic signals in open hole and again after cementing casing. Observations and conclusions drawn from this experiment are included. Introduction Since Nov., 1960, approximately 250 cement bond logs have been run in fields along the Cedar Creek anticline. This structural feature is located in the western portion of the Williston basin, primarily in southeastern Montana. The average interval covered by most of the bond logs was from 5,000 ft to a total depth of 8,500 to 9,000 ft. The upper half of this interval includes predominantly shales with some sands and evaporite sections, most of which have acoustic velocities slower than steel. The lower half of the interval contains mostly dense carbonates which frequently exhibit velocities faster than steel, causing formation signal arrivals ahead of the casing signals and, thus, complicating bond interpretation. A continuing effort has been made to assess the variables affecting the log. Logging tools have undergone several important changes since they were first introduced, further complicating the interpretation problem. Therefore, this investigation incorporates all of the environmental and design variables observed in an attempt to assess both the positive and negative aspects of cement bond log interpretation. The discussion was divided into four categories for clarity of presentation. However, they are all closely interrelated and, of necessity, overlap somewhat. One well, Unit 32–14-17, was logged extensively with as many tools as practicable including a log which photographed received signals both in open hole and in casing. This well is referred to as the experimental well and is used as an example throughout the report. The scope pictures are used qualitatively because it was not possible to relate absolute amplitudes with the commercial tools. It is assumed that the reader is familiar with the basic concepts involved in acoustical cement bond logging as outlined in several earlier papers on the subject. The terms "bond" and "bonding", as referred to in this report, allude to mechanical as well as adhesive bonding. Logging Tool Designs To properly interpret any bond log, it is important to understand the type of tool being run. There are a number of different types available to the industry, each with its merits and disadvantages. Any two of these tools run under exactly the same conditions could have entirely different interpretations if the tool characteristics were not known by the observer. The tools discussed in this report. as well as all tools in use today, can be placed in two broad categories:surface recording instruments bring the entire acoustical signal to the surface for display on a scope in addition to recording amplitude on a standard log grid; anddown-hole recording instruments do not employ a scope and bring only those portions of the signal to the surface which are thought to pertain to bonding, namely, the amplitude of a specific wave peak and, in some cases, the time this signal arrives. To better understand some of the terms used as specifically related to cement bond logging, a typical acoustic wave train as seen in unsupported casing is shown in Fig. 1. Regardless of whether the tool is surface recording or down-hole recording, the acoustic signal it "sees" is the same. Amplitude is measured as either a positive or negative vertical deflection from the base line, and time is measured horizontally along the base line. Usually, only certain specific amplitude peaks of the wave train are measured. This is accomplished by electronic gate circuits, labeled 1, 2, 3 and 4, to represent the various gate widths employed by the four commercially available logs studied in this report. JPT P. 17^