Abstract The technology of the oil industry today is such that more rigid requirements are placed upon cement isolation in wells. In addition, recent advances in well evaluation and completion technology permit completions to be made in zones that would have been considered uneconomical a few years ago. With longer payouts, it is even more important to insure the best possible completion. Many completions are made in formations that depend upon natural fractures, vugs, or induced fractures for commercial production rates. Maximum production under these conditions requires cement isolation before completion attempts to prevent damage resulting from squeezing cement, especially after any well treatment. Acoustic cement bond logging is one of the tools in well completion technology that can be used to insure the best possible completion by insuring isolation of all zones before a completion attempt is made. It shows the degree of isolation. Under many conditions the cost of the log is small in comparison with squeezing, reperforating, refracturing, decreased production, or even loss of a well. Field examples illustrate a number of cases where considerable extra completion expense arose because the information from the bond log was not used. Basic bond log interpretation is included in the Appendix. Introduction Many conditions have placed more rigid requirements upon the effectiveness of cement isolation behind casing. Deeper drilling, with the accompanying higher pressure and temperature, has resulted in higher pressure differentials on the cemented interval. This has required more bonded interval for effective isolation. The increased application of high-volume, high-rate well treatments requires the cemented interval to withstand high pressure differentials. Secondary recovery of all types also requires complete isolation behind casing, both from the standpoint of cost of injected fluids and efficiency of operation. Often, a production test is the most economical evaluation of cement isolation. When there are no other permeable zones near the completion interval, the odds are that isolation is sufficient. In formations of high permeability, when a production test shows channeling from another zone, cement squeezing and reperforating will not greatly reduce productivity. However, there are many other situations in which the economics are such that complete knowledge of the degree of cement isolation should be obtained before perforating. Acoustic Cement Bond Log Application There are many variables that control the effectiveness of the cement in a well depth, temperature, hole size, additives, contamination, type of cement, type of cement flow, etc. Even with the best available cementing program, channels or some other type of unbonded section can exist in critical intervals. Because of these uncertainties, the usage of the acoustic cement bond log has increased greatly in the past few years. The acoustic wave in a cased borehole consists of all arrivals along any coupled path between transmitter and receiver. A recording of this entire acoustic wave, properly interpreted and used (see Appendix), can supply the information needed to design the most economic completion procedure. One possible presentation is the intensity-time recording where dark and light streaks represent the positive and negative half cycles of the acoustic wave. Amplitude is shown by the darkness or lightness of the streaks. The position of the streaks from left to right denotes increasing arrival time. Fig. 1 is an intensity-time recording presentation on a bond log run in 4-in. liner. Only 5 ft of good acoustic bond is indicated above the interval to be perforated. The reservoir pressure at the time of completion had decreased to approximately 5,500 from 9,000 psi originally. Because of high pressure differentials, it was recommended that the liner be squeezed above. However, this was not done, and the well came in producing gas at the rate of 4 to 5 MMcf/D. After several months of production, the well suddenly died. During cleanout operations, sand, shale and cement were recovered from the well. Before the workover was completed, the liner collapsed, and the well had to be plugged. Above the main pay sand there were thin sand stringers at original reservoir pressure that were the probable cause of the liner collapse. In this case, a squeeze before perforating for production and even another bond log run would have been economical. Figs. 2 and 3 show this same acoustic intensity-time presentation on a bond log run in a well that was intended as a triple completion. JPT P. 811ˆ