Tracer-Placement Techniques for Improved Radioactive-Tracer Logging

Summary. The accuracy of a radioactive-tracer log depends on the tracer rapidly mixing with the wellbore fluid and traveling with the same velocity as the wellbore fluid. In addition, log quality depends on the ability to measure the tracer transit time from the gamma ray detector responses accurately; the tracer must maintain a sharp peak, or sharp leading edge, for reliable transit-time measurement. We conducted experiments in a model wellbore to determine the factors affecting tracer slug distribution and the responses at tracer detectors resulting from these factors. Tracer-placement parameters studied included tracer shot velocity, shot duration, wellbore flow rate, and tracer viscosity. When wellbore flow was laminar, the tracer dispersion could be predicted with a theoretical model and the experimental results compared well with the theory. As would be expected, optimal results were obtained when the bulk of the tracer was placed in the high-velocity, central region of the wellbore. In turbulent flow, a fairly low ejection rate and large shot size most consistently yielded sharp slugs that were maintained at both detectors. Introduction Radioactive-tracer logs are commonly used to measure injection profiles by timing the passage of a slug of radioactive tracer between two gamma ray detectors (the velocity-shot log). Tracer log accuracy depends on the tracer rapidly mixing with the wellbore fluid and traveling with the same velocity as the wellbore fluid. In addition, log quality depends on the ability to measure the tracer transit time from the gamma ray detector responses accurately; the tracer must maintain a sharp peak, or sharp leading edge, for reliable transit-time measurement. A new tracer logging method, the two-pulse log, also depends on the maintenance of sharp peaks of the tracer slugs. The accuracy of a radioactive-tracer log thus depends on the way in which the tracer is introduced into the wellbore and how it mixes and disperses in the wellbore fluid. We conducted experiments in a model wellbore to determine the factors affecting tracer slug distribution and the responses at tracer detectors resulting from these factors. The tracer-placement parameters studied included tracer shot velocity, shot duration, and wellbore flow rate. In addition, the viscosity of the tracer was varied to determine if this would decrease the dispersion of the tracer slug. A blue dye was used as tracer, and two detectors downstream of the injection point measured tracer concentration on the basis of attenuation of light passing through the clear pipeline. Tracer slugs were also videotaped to determine their distribution across the wellbore. In the analysis of detector responses, the relative sharpness of tracer slugs was quantified by normalizing the width of the tracer slugs with the peak heights and measuring the spread of the tracer as it passed from the first detector to the second. When wellbore flow was laminar, the tracer dispersion could be predicted with a theoretical model and the experimental results compared well with the theory. As would be expected, optimal results were obtained when the bulk of the tracer was placed in the high-velocity, central region of the wellbore. In turbulent flow, a fairly low ejection rate and large shot size most consistently yielded sharp slugs that were maintained at both detectors. A high-viscosity tracer also performed better than a tracer with about the same viscosity as the wellbore fluid. According to these results, it is clear that radioactive-tracer log quality can be significantly improved by using the newer logging tools that allow for careful control of shot duration. With a few test shots in the upper part of the wellbore, optimum tracer placement techniques for a given well can be determined. Radioactive-tracer logging is common for measurement of flow profiles in wells, primarily water-injection wells. Two logging methods are prevalent: the tracer-loss and velocity-shot methods. In the tracer-loss method, a single slug of tracer is ejected into the wellbore above all zones of fluid loss; the amount of tracer in the wellbore is then measured as a function of depth by repeatedly passing a gamma ray detector through the tracer slug as it moves down the wellbore. The amount of tracer remaining in the wellbore at any depth is assumed to be proportional to the flow rate in the wellbore at that depth. A velocity-shot log consists of measuring the transit time of a tracer slug passing between two points (usually between two gamma ray detectors, but sometimes between the tracer ejector and one gamma ray detector). From the transit time, the wellbore fluid velocity is calculated. By repeating such velocity measurements at numerous locations in the well, the flow profile can be constructed. Of the two common radioactive -tracer-logging techniques, Hill and Solares have shown that for virtually any well conditions, the velocity-shot method will provide the more accurate flow profile. A proposed new logging method, the two-pulse log, appears to have greater accuracy than the tracer-loss log and advantages over the velocity-shot log in wells with varying cross-sectional areas. To run a two-pulse log, two slugs of tracer are ejected a known distance apart above all fluid exits, either by firing two ejectors simultaneously or by sequentially ejecting two slugs with one ejector and then logging through the slugs with a gamma ray detector to determine the initial spacing between the pulses. The log is then run in the same manner as a tracer-loss log. The distance between the peaks of the tracer slugs are measured as a function of depth as they move down the wellbore by repeatedly logging through the tracer slugs with a gamma ray detector. As fluid is lost to the formation, the two pulses of tracer will move closer together; the distance between the pulses is proportional to the wellbore flow rate. Because of their greater accuracy compared with a tracer-loss log, the effects of tracer placement on the velocity-shot and two-pulse logs are the focal points of this paper. A velocity-shot log is analyzed by measuring the transit time between two detectors, At, and then calculating fluid velocity as (1) where v is velocity and L is the spacing between detectors. The velocity calculated may be either the maximum, center-line velocity or the average velocity, depending on whether the flow is laminar or turbulent and on the manner by which the transit time is measured. JPT P. 1484^