A Progress Report on Cement Bond Logging

Abstract

PUTMAN, LLOYD, OIL WELL PERFORATORS, INC., CASPER, WYO. MEMBER AIME Abstract During the past year notable progress has been made in acoustic cement bond tool design as well as improvements in operating techniques. These developments have been a direct result of suggestions and challenges presented in previous papers by such authors as Pickett, Bade, Walker, Anderson, Riddle and others. These writers noted the many variables and inherent difficulties involved in running and interpreting cement bond logs. During the development process of combating these problems, more than 100 cement bond logs were run in the Wyoming and Montana areas. On the majority of these surveys a development engineer was present and prepared to make design changes under field operating conditions on a trial and error basis. In each case the well operator allowed extra rig time or else a portable mast was used. In addition to commercial logs, an 8,900-ft test well on the Cedar Creek anticline of Montana areas used to perfect techniques of calibration and make other more difficult design and circuit changes. Illustrations and field examples are presented and discussed to support conclusions reached. Introduction Since late 1962 a field development program involving more than 100 acoustic cement bond logs has been underway in fields of Wyoming and eastern Montana. This effort has been toward developing a practical and routinely usable logging system which incorporates a maximum number of desirable qualities mentioned in previous papers. With the vast amount of actual information available through writers such as Walker, Bade, Anderson, Pickett, Riddle and others, it is not surprising that a vigorous field development program would produce a refined logging instrument. Results have shown that with proper tool design many of the mechanical and electrical problems now existing can be overcome. Also, it appears evident that evaluation of the effectiveness of cementing under most of the known conditions is now reasonably reliable. Quite unique in this development program was the presence of a development engineer on commercial jobs who was prepared to make design changes under field operating conditions. On many occasions numerous runs were made on a trial and error basis to achieve desired results. The advantage of this type arrangement is apparent because the engineer could run a survey, come out of the hole and make simple or major modifications, go back and re-run the log and repeat as many times as necessary until desired adjustments and responses were effected. In addition to commercial logs run, an 8,900-ft test well on the Cedar Creek anticline of Montana was used to perfect calibration techniques and make more difficult design and circuit changes. Progress is reported concerningfast formation response;use of shear waves in bond logging;logging boreholes containing oil and water; andlog calibration suggestions. The fundamentals of acoustic cement bond logging are by this time well known. For the purpose of this paper, bond logging implies an attempt to correlate acoustic coupling of casing to cement and formation to a hydraulic bond. Although no attempt is made in this paper to ascertain the indicated degree of bond to actual hydraulic seal by production testing, work is progressing satisfactorily along this line. For bond logging to be completely successful this work is gravely needed to properly determine when remedial action should be initiated from the bond log. New steps and refinements in techniques toward achieving this goal are discussed individually. Fast Formation Response The term "fast" or "early formation arrival" has become a popular term, and normally is used to describe an interpretation problem that has existed in the past when logging hard formations. Simply, it implies that a portion of the energy from an acoustic transmitter may be transmitted through casing, cement and formation and arrive at a receiver simultaneously or faster than energy from the same source can travel a physically shorter path by way of the casing and cement annulus (illustrated in Fig. 1, Path B over A).Presently the method for detecting this situation is to run a single receiver-time curve simultaneously with the bond log. For some time this method has been accepted as a satisfactory solution to the problem, however, recent study indicates a more direct solution. JPT P. 1117ˆ