New logging-while-drilling (LWD) nuclear technology improves neutron log quality.

LWD density and neutron tools produce optimum logs when run in an in-gauge borehole with minimal (or no) standoff from the formation. In this scenario, the presence of borehole fluid between the formation and detectors is minimized. This is important because the borehole fluid attenuates the gamma rays and neutrons traveling from the formation to the tool's detectors, negatively impacting the neutron porosity and density measurements. Simply put, the greater the standoff, the greater the influence on the measurement accuracy. Standoff effects are especially critical in out-of-gauge boreholes and when LWD tools are run behind steerable mud motors in extended-reach or horizontal applications.
In wireline nuclear applications, this challenge is overcome by pushing the tools against the side of the wellbore to follow the enlarged boreholes' irregular profiles. Unfortunately, this conveyance technique is not available in LWD tools. In LWD tools, the nuclear detectors are mounted in drill collars with fixed ODs that, while rotating, are unable to maintain constant contact with the borehole wall. Drilling occurs under very dynamic conditions, and the position of the bottomhole assembly changes rapidly, causing LWD tools to bounce laterally off the borehole wall in an unpredictable manner. Under these conditions, the standoff distance between the LWD tool and borehole wall changes in a similarly rapid and unpredictable manner. Such changes in the standoff adversely effect the accuracy of the density log.
Different methods are used to compensate for the LWD tool's standoff and improve the accuracy of real-time nuclear logs. The most common method divides the borehole into four sectors, and prior to drilling, a single sector is chosen to represent the formation. Unfortunately, a variety of drilling conditions could prevent the preselected (low side) sector from providing a representative log.
A new LWD tool suite and logging method have been developed to measure and reduce the impact of borehole size and standoff on LWD density logs. Standoff-based acquisition, or binning, simultaneously measures standoff and density at high sampling rates and compensates the log for a range of standoffs. In this binning process, nuclear data are acquired and stored in bins designated by a specific standoff range in which the data were measured. This method provides a means to optimize the log response during rotation.
The new LWD density tool houses three ultrasonic transducers, one of which is mounted in line with the density detectors to measure its standoff. It forms part of a three-axis, ultrasonic caliper that measures borehole size. The caliper is used to correct the neutron porosity log for borehole size and standoff.
The technique has been documented in laboratory and downhole formations. Comparisons between LWD and wireline logs provide a measure of the accuracy of corrections while sliding and rotating. Examples include quality control logs based on the time spent in each bin, bin weighting, the data frequency distribution (by bin) and the different filters available to enhance log resolution.
The tool suite
The Caliper Corrected Neutron (CCN) and Optimized Rotational Density (ORD) tools are compatible with probe and collar-based architectures.
The CCN tool represents a significant change from previous LWD neutron tools. The detector configuration has been modified to improve count rates and borehole sensitivity, and acquisitions have been increased from two per minute to 12 per minute (one acquisition every 5 seconds). The caliper measurement from the corresponding density tool can be used to provide a more accurate borehole size correction.
Because the source and detectors are subwall-mounted and closer to the borehole wall, the necessary borehole corrections are relatively small. However, the smaller borehole corrections can be partially offset by standoff effects. Although standoff effects are generally small when logging in a rotary mode, they will increase in a sliding mode whenever the detectors are not oriented to the low side of the hole. Under these conditions, knowledge of the tool orientation (relative to the low side of the hole) and the three-axis caliper measurement from the density tool are used to determine the proper standoff correction required from the acquisition software.
The ORD tool includes three acoustic transducers mounted just below the long detector. Transducer No. 1 is in line with the density detectors, and the other two are circumferentially oriented 120° to either side. Each transducer records distance to the borehole wall 200 times per second (this is equivalent to one measurement of standoff every 6° while rotating the drillstring at 200 rpm). All three measurements are combined and provide a measurement of borehole size or caliper. The measurements from transducer No. 1 also are used in the density acquisition. As with the neutron tool, acquisition rates for the density tool are 12 per minute.
Hard facing on the exterior of the subs, wear pads and new density window materials provide resistance to wear, and new packaging methods for the electronics and detectors offer added reliability.
Intelligent acquisition method
Density logs are susceptible to standoff-based errors, and a variety of methods are used to minimize these errors in LWD logs. The most common standoff correction method for real-time logs divides the borehole into predefined, angular sectors and produces a separate log from each (Figure 1a). Using this method, a single sector (typically from the low side) is preselected to represent the formation log. This method assumes that data from the chosen sector is good and most representative. Unfortunately, this is not always the case given the rapid, and often violent, toolstring rotation.
However, the new CCN/ORD tool suite employs a new acquisition technique to eliminate standoff-based errors (Figure 1b). During acquisition, the density tool's downhole microprocessor correlates counts recorded at each detector to the standoff measurements made by transducer No. 1. The microprocessor bins each acquisition according to standoff (Table 1).
This method integrates a direct measurement of standoff into the density acquisition. As the tool rotates around the borehole, density measurements with the least amount of standoff are separated from those with excessive standoff. In addition, a short- and long-spaced density, compensated density and correction are determined and stored for each bin. At the end of the acquisition period, adaptive weighting is applied to the binned data, and a compensated density is calculated using data from all five bins. The adaptive weighting ensures statistical error and standoff are minimized in order to obtain an optimal density value.
Table 2 illustrates a typical acquisition. Although the density from Bin 1 has the least amount of standoff, it also represents only one-tenth of the acquisition period. The statistical error would be too high to calculate a density from this bin alone, so it must be combined with the other bins. In this case, Bin 2 is used with equal weight. Bin 3 is also used, but as one might expect with declining weight because of the greater amount of standoff. Typically, Bins 4 and 5 are rejected as a result of the large standoff associated with these measurements. Only when the bulk of the acquisition data lies within Bins 4 and 5 are they used for the final calculation of density. When this occurs, quality flags identify a questionable acquisition.
Figure 2 offers a comparison between the wireline caliper, neutron porosity and density and the CCN, ORD and the ORD tool's three-axis caliper. The LWD log was acquired while rotating behind a steerable assembly. The borehole inclination was nearly 60°, and the mud weight was 13.2 ppg.
Gamma ray and caliper are plotted in Track 1. The LWD caliper has not been corrected for mud weight, temperature and pressure. As a result, it shows a slight discrepancy with the wireline caliper (about 0.25 to 0.30in. larger). The wireline and LWD neutron porosity are plotted in Track 2. The wireline and LWD density are plotted in Track 3. The time spent in each bin is plotted in Track 4 as a quality control indicator. This provides a snapshot of the distribution of data within the five bins through the interval logged.
The excellent agreement between the wireline and LWD logs in the reservoir rocks demonstrates the improvements in the neutron response and the robustness of the density tools' intelligent acquisition method despite considerable standoff as indicated by the time in each bin plot in Track 4. A new filtering routine more appropriately handles the transformation from unevenly sampled to evenly sampled data while retaining the true resolution of the neutron porosity and density measurements. This routine also minimizes the filtering of real geologic features and eliminates processing artifacts.
This method removes all of the assumptions required by other LWD acquisition methods and compensates for poor borehole conditions and drillstring dynamics. Several quality control outputs permit evaluation of data quality as it relates to standoff ensuring that the best density measurements are preferentially selected despite adverse logging conditions.