PRCI

Recent Results

Salt Caverns in Gas Storage Service

In a project cofounded with the Gas Storage Technology Consortium, RESPEC Engineering (Rapid City, SD) is examining the issues surrounding the use of brine string tubing temperature log anomalies as a possible mechanical integrity test (MIT) for salt caverns in gas storage service. Current MIT options can be quite costly and may require the cavern to be removed from service for an extended period. In simplest terms, any gas leak somewhere along the brine string would cause local Joule-Thomson cooling, which would not only indicate the location of the leak, but also create a temperature profile unique to that leak that would also be proportional to the size of the leak. Cold spots can be readily measured by temperature probes in the brine string.

Key Results

RESPEC’s evaluation was two-fold. In a prior project, they catalogued the failure modes of salt caverns, and the present phase examines whether this temperature log diagnostic is sufficiently robust to be relied upon for use as a MIT, given the possibilities of potentially catastrophic “false positives” (test indicates good well integrity when there actually a leak), as well as the cost of false negatives (test indicates there a leak that does not actually exist). While that core question was not resolved, RESPEC’s CFD modeling of leaks as a function of casing types indicated that both turbulent and laminar flow conditions in the casing, at different annulus gap sizes, would still allow cold spots to be detected. Therefore, temperature logging for leak detection is not limited by the casing combinations found in common well completions.

This work will be continued in a 2010 PRCI project, which (via additional modeling) will determine the upper bound for the magnitude of the temperature anomaly that can be created by different leak rates. In addition, the project will provide an illustration of the expected temperature distribution and flow characteristics in a well under MIT conditions. The models will simulate wells that have a cold spot. The results of the modeling effort will include the location and intensity of convection cells that develop within the gas-filled annulus. That will contribute to a better understanding of the thermodynamics of the cased well with a leaking brine string.  If the modeling results are found to converge, a field test will be conducted to validate the model further, that model to provide the granularity necessary to determine whether this approach can be a substitute for current MIT methods.

Return to Recent Results Front Page