The deteriorating condition of the American Petroleum Institute (API) gamma ray calibration facility on the University of Houston campus was high-lighted in two papers presented at the 2017 Annual Logging Symposium. Now that awareness of the API facility's rapidly approaching end of life has been raised, the search for a suitable replacement should include a variety of ideas. The ultimate solution must be readily accessible to service companies regardless of size or market share, and it must be applicable to wireline and logging-while-drilling (LWD) tools, which had not been thought of when the API facility was designed. Construction of a new, physical replacement for the API facility that would accommodate both types of gamma ray instruments is an obvious solution. However, there is little interest among the major service providers and oil companies to embark on such a complex and expensive undertaking in the current business climate.
Monte Carlo modeling of nuclear logging tool responses has become a widely accepted and cost-effective practice during the previous 40 years. The future replacement for the API calibration facility will probably involve Monte Carlo modeling because of these features. For this to be a viable option, a normalization to measured data obtained from a test pit, tank, or fixture is necessary to obtain accurate results.
This paper proposes a new method using calibrated Monte Carlo models to obtain counting rate-to-API unit sensitivity factors for new hardware de-signs by modeling their responses to thorium blanket calibrators. Most, if not all, service providers have thorium blanket field calibrators characterized with existing natural gamma ray tools calibrated at the API calibration facility. The new technique, which can be applied to wireline and LWD tools, is advantageous because it uses existing field equipment to calibrate the Monte Carlo models instead of bulky, fluid-filled tanks or other constructions, such as man-made or natural rock formations. However, Monte Carlo techniques not previously used for well logging applications are necessary to accurately simulate tool responses to thorium blanket calibrators. Results for multiple tool configurations are presented that show the new technique produces results, which are typically accurate to within a few tenths of a percent. The paper also includes sufficient details to enable anyone with modest nuclear modeling experience to apply the technique to their equipment.
