Gamma-Ray-Tool Characterization Avoids Traditional Limitations
In this paper, the authors discuss the characterization process for GR tools and how they behave in boreholes different from the one used in the University of Houston (UH) GR characterization pit.
A major issue with gamma-ray (GR) logs is that the API definition is valid only if the tool is run in a 4.89‑in. borehole filled with fresh water. Although all GR tools are intended to provide the same results for such a well, in the field there is no single one-size-fits-all concept. In this paper, the authors discuss the characterization process for GR tools and how they behave in boreholes different from the one used in the University of Houston (UH) GR characterization pit. Proposals for developing correction strategies so that GR logs become quantitative logs, rather than the qualitative logs of the past, also are outlined.
One of the issues inherent in ensuring repeatability of GR logs starts with tool characterization. For decades, no common industry standard existed that defined the scale for measuring radioactivity. The chaos created by different tools providing radioactivity in different units in the same environments ended with the creation of the API unit in 1959. The definition of the unit was based on work done with the UH GR characterization pit. This specific pit has a 4.89-in. borehole and a casing to ensure stability within the surrounding synthetic formation.
One downside of the API definition is that different combinations of radioactive isotopes, environmental conditions, and tool designs may yield comparable count rates. UH GR pit characterizations of GR tools lose their validity as the wellbore diverges from this previously described specific borehole size and environment. Thus, environmental corrections are needed to bring readings to a reference borehole used for tool characterization.
Another issue with the UH GR-pit-based characterizations became evident with the emergence of logging-while-drilling (LWD) GR tools. The LWD tools were supposed to provide natural GR values comparable with those of wireline tools. Recommended practices for wireline tools did not have a provision for calibration of LWD GR tools that did not fit into the UH GR pit. Consequently, characterizations of LWD tools to provide results comparable with those of wireline tools became a topic of interest.
Although the UH GR characterization pit still exists, it suffers from many issues. The facility is in a state of disrepair, with significant corrosion. In addition, its activity level is decreasing because of the presence of short-lived Ra-226 as one of the main ingredients in the high-activity zone. The radionuclide distribution is not even, causing gamma readings to change with the height, and rock dimensions do not correspond to those of an infinite rock environment.
No recommended practices exist for situations in which the tool is run in boreholes larger than 4.89 in. that are filled with water or heavier muds. Environmental corrections are employed for cases in which the borehole size is different from the reference borehole size used for characterizing the tool. This brings up the question of the reference borehole size for such corrections. This is undefined, and has significant implications for the delivered GR values in the API unit.
How To Define API for LWD Tools and Reference Borehole Definition
One of the important aspects of LWD GR tool characterization is the need for a defined characterization procedure that ensures that the specific LWD tool provides GR values comparable with those provided by wireline tools in the same rock, regardless of the borehole size. The procedures employed by different companies deliver tool characterizations that result in different LWD GR API logs.
LWD tools come in different dimensions, with 4¾-, 6¾-, and 9⅞-in. tool sizes being especially popular. The reference borehole sizes for those tools are of the utmost importance because the tool should provide the same GR reading as a wireline tool provides for the same rock under reference borehole conditions. One additional step that the authors propose for LWD GR tools is that they have their reference boreholes set up in such a way that LWD tools provide the same API value for the given rock that the wireline tool would provide when corrected with the borehole size and mud-weight corrections based on 4.89-in. boreholes filled with water. The chosen reference borehole size for a given LWD GR tool should be close to the tool size. It is favorable to choose the smallest borehole size possible for the given tool as the reference borehole.
Environmental corrections for natural GR tools are currently based on comprehensive computer-based Monte Carlo radiation transport simulations, verified against test runs in reference rocks performed in the laboratory. To run nuclear simulations of GR interactions in the rock formation and the logging tool, detailed knowledge of interaction cross sections, material properties, and complex natural decay chains is mandatory. As mentioned previously, there is no reference borehole definition for LWD GR characterizations and environmental correction. The only reference borehole definition exists for the wireline tool characterizations, but none exist for borehole size and mud-weight corrections.
Fig. 1 presents various GR logs in a 9.2‑ppg mud-filled 6-in. borehole of a test wellsite. The well has been logged using a 3⅝-in. decentered wireline tool and a probe-type LWD GR tool that can fit into this size well. While the wireline tool has been characterized and calibrated in accordance with API Recommended Practice 33 based on the UH pit, the LWD GR tool has been characterized on the basis of a 6-in. water-filled reference borehole. Therefore, LWD gamma readings did not need borehole-size corrections, but a small mud-weight correction of 1.027 was introduced to account for the presence of 9.2-ppg mud in the system.
The left panel of the log in Fig. 1 shows the wireline GR corrected using borehole size and mud-weight corrections made on the basis of different reference borehole sizes. The green curve in this panel shows the API values if the correction scheme is based on a 4.89-in. water-filled borehole, as in the UH pit. The blue, orange, and red curves are logs corrected using 6-, 8½-, and 9.875-in. reference boreholes. Using different reference boreholes for corrections results in significant shifts in the corrected GR logs.
The middle panel presents the corresponding GR log of the 4.75-in. LWD run in the same test well. API characterization of the tool is based on a water-filled 6-in. borehole. Again, corrections are applied. The corrections in Fig. 1 are based on a 6-in. (blue) reference borehole, as well as 8½-in. (orange) and 10-in. (red) renormalizations corresponding to scale factors of 0.867 and 0.792, respectively. Furthermore, a mud-weight correction of 1.015 was introduced. As with the wireline tool, a significant drop in the corrected GR values is evident with increasing reference borehole size.
The wireline GR and LWD GR should match when they are both in the reference boreholes. The LWD log is already in the reference borehole. It was characterized in a 6-in. water-filled wellbore, and corrections are based on a 6-in. water-filled borehole. Except for the mud weight, the correction factor corresponds to unity. Nonetheless, the wireline tool is not in the reference borehole; thus, it should be corrected. The corrections should have a reference borehole the same as the API characterization borehole. For wireline tools, that is a water-filled 4.89-in. wellbore. The reading within the 6-in. borehole is corrected to the 4.89-in. reference borehole value using 4.89-in. reference-hole-based corrections, which is represented by the green curve in the left panel in Fig. 1. After applying the correction factor of 1.056 and the mandatory mud-weight correction, both LWD and wireline curves overlap, as visible in the right panel of Fig. 1.
Work Flow for Consistent Corrections
How can existing algorithms be used to bring apparent API values and corrections to the same level, so that resulting corrected GR logs will be quantitatively meaningful? For this purpose, the authors provide two flow charts in the complete paper.
One of these work flows is specific to wireline cases. The API characterization for wireline tools should be based on the water-filled 4.89-in. UH test pit. The first task in obtaining the corrected GR log is to subtract the borehole-fluid potassium counts, if any potassium is present in the borehole fluid. Next, a correction factor for a water-filled 4.89-in. borehole is computed through a service provider’s correction algorithm. This correction factor is then used to obtain a normalized correction algorithm by dividing the service provider’s correction algorithm. This normalized correction algorithm is then used with the apparent GR log to obtain the corrected GR log.
The other work flow is for the LWD GR logs. This has an additional step. Because no standard reference borehole size for LWD API characterizations exists, this should be obtained from the service provider so that specific borehole size and borehole fluid can be used to obtain the correction factor using the service provider correction algorithm. That same algorithm is then normalized. The apparent GR logs are then corrected using this normalized correction algorithm.
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 191717, “Is API Enough for Gamma Ray Logs, or Do We Need More?” by Feyzi Inanc, SPE, and Andreas Vogt, SPE, Baker Hughes, a GE Company, prepared for the 2018 SPE Annual Technical Conference and Exhibition, Dallas, 24–26 September. The paper has not been peer reviewed.