Harsh Environment Wells

Through-tubing hostile-environment acoustic logging tool for cased holes

Positive results of a new small-diameter, cased-hole hostile environment wireline sonic tool is proved by two case history applications.

Lucio N. Tello, Daniel L. Long and Thomas J. Blankinship, Computalog, Precision Drilling; and Marney N. Pietrobon and Robert L. Horine, BP America

The drilling environment encountered in modern wells, especially gas wells, frequently includes higher temperatures, higher pressures and smaller diameters than standard wells. Further, specialized borehole fluids or air may be used to preserve borehole and formation integrity. These conditions frequently limit or even preclude use of conventional acoustic logging tools. High rig costs may also result in the elimination of all, or part of the traditional openhole logging program.

Described here is the development of a new, slim acoustic logging tool for cased-hole cement-bond evaluation and more complete formation evaluation. Two case histories demonstrate the new tool's versatility.

The new slim, hostile-environment wireline acoustic logging tool was developed for the cased-hole market. The small diameter, 1-11/16 in., allows tool operation in casing or tubing as small as 2 in., and also in standard drill pipe. The tool is rated for operation at temperatures and pressures up to 420°F and 20,000 psi.

A full processor-controlled, digital acquisition system, in combination with a carefully designed arrangement of acoustic transmitters and detectors, provides a simultaneous high-quality cement-bond log, and monopole compressional slowness or radial channel identification. The addition of pulsed-neutron measurements allows for a more complete formation-evaluation analysis.

Case histories and test-well studies demonstrate the full range of the tool's measurement capabilities for formation evaluation, seismic correlation, cement evaluation, and stuck-pipe analysis in a variety of borehole environments, and in casing sizes ranging from 2-7/8 in. to 9-5/8 in. In one example, twin wells were drilled in an onshore gas field, one using air, the other conventional borehole fluid.

For this type of well, air drilling is a more efficient and lower-cost construction method, but it makes openhole acoustic logging impossible. Logs were run in the fluid-filled open borehole and also in the air-drilled well after setting casing. The cased-hole data proved superior to the openhole data, possibly due to formation alteration by the borehole fluid in the openhole, and provided more accurate synthetic seismograms and seismic ties.

In another example, the BHA was stuck in the well and the through-tubing acoustic tool was run through drill pipe to identify stuck pipe intervals. Having this information was critical for planning the successful fishing operation because it enabled decisions regarding wash-pipe length, free-point indication and backoff, and allowed estimates of the remaining downtime.

In situations where openhole logging is either impossible or undesirable, cased-hole acoustic logging can provide high-quality data at a lower cost.

INTRODUCTION

Compressional slowness is important in many applications, including porosity estimation, lithology identification and seismic correlation. While compressional slowness is generally obtained in openhole, it is also possible to extract acoustic velocities in cased holes by applying a variety of advanced processing algorithms to the data obtained by modern array-type tools.^1,2

Although dipole array tools typically perform better than monopole arrays in casing,³ their larger diameters make them unsuitable for use in smaller casing sizes. Increasingly, new gas fields require deeper, hotter and smaller diameter boreholes which, combined with high rig cost, limit, and in some cases, eliminate, the possibility of openhole logging. While formation evaluation is important in cased wells, equally important is the cement to casing bond evaluation, and cement to formation bonding indication, as well as channel identification. The Slim Monopole Array (SMA) tool was specifically developed to provide cement-bond evaluation, as well as formation compressional slowness in casing diameters ranging from 2-7/8 in. to 9-5/8 in.

Two case histories demonstrate the versatility of this new tool. The first is an example of cased-hole formation evaluation. It involves two adjacent wells in which data in one was obtained in a fluid-filled openhole – in the other, through casing in an air-drilled well. The example demonstrates that acoustic data obtained in the cased hole, combined with pulsed neutron decay (PND) data, are sufficient for formation interpretation and analysis.

The second example describes use of the SMA tool in a fishing operation involving stuck pipe. In this case, the small tool diameter allowed it to log through drill pipe and provide correlating information to the free-point-indicator tool.⁴

Two additional studies noted herein, and described in detail in the SPWLA paper referenced in the Acknowledgment, present: 1) a cement-bond evaluation in a test well with channeling behind casing, ranging in size from 4-1/2 in. to 9-5/8 in., in which circumferential resolution results are compared to the standard 2-3/4 -in. Sector Bond Tool,⁵ and 2) a study of stuck-pipe logs using an example West Texas well.

TOOL DESCRIPTION

Fig. 1. Slim Monopole Array Tool (SMA), a; and Slim Sector Bond tool (SSB), b. Difference between the two is receiver configuration contained in the middle section.

The SMA is a 1-11/16 -in.-diameter, hostile-environment, wireline sonic tool. The slim-hole design allows logging in tubing as small as 2-in. ID, and it can operate at temperatures and pressures up to 420°F and 20,000 psi. Other features include specially designed gas-proof metal-canned transmitter and detectors, a 16-bit downhole digitizer that provides greater dynamic range than previous tools, and a downhole programmable variable gain amplifier for optimum signal to noise ratio.

The tool comprises an upper section containing an electronics cartridge, and a middle section consisting of an array of eight monopole detectors spaced 3.0, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5 and 8.0 ft from a single monopole transmitter located at the lower section of the tool, Fig. 1a. For radial cement and channel identification, the detector array (middle section) can be exchanged for an array of eight detectors, six of which are arranged as a six-sector circumferential detector having a 60° spacing. In this latter configuration, called the Slim Sector Bond (SSB) tool, the other two detectors are used for the 3- and 5-ft T-R spaced CBL and VDL applications, Fig. 1b.

The downhole acquisition system used to retrieve the sonic data incorporates state-of-the-art electronics technology to preserve signal integrity. The programmable gain amplifier has a wide bandwidth; thus, no filters are applied to the received signals. This is of particular importance, since downhole filters can alter the frequency content of the signals that are essential to separate formation and the casing components of the wave train.

COMPARING SEISMIC CORRELATION AND FORMATION POROSITY

In the first case history example, two adjacent wells were drilled in the same field, less than two miles apart. Well 1 was drilled with water-based mud and openhole logs were run, including the sonic delta-T. Well 2 was air-drilled, cased and cemented. Acoustic and pulsed neutron logs were run at a later date when completion fluid filled the borehole.

Synthetic seismograms generated from each well are shown overlain on the seismic data in Fig. 2. Both synthetics show good correlation to the seismic above the Marker Bed. However, below this marker, the synthetic seismogram from Well 1 begins to stretch and become progressively less well correlated to the seismic data. The synthetic seismogram from Well 2 remains well correlated throughout the reservoir section. The signal-to-noise ratio in the match between the synthetic and the seismic, was 60% higher in the well which was logged through casing.

Fig. 2. Seismic section with synthetic seismograms generated in Wells 1 and 2 overlaid. Compared to Well 2 (cased hole), the correlation quality in Well 1 (openhole) decreases with depth.

PROCESSING SMA WAVEFORMS FOR COMPRESSIONAL SLOWNESS

Semblance waveform correlation techniques are used to derive compressional slowness from monopole waveforms recorded in a cased-hole environment.

To get a general sense of what the data can yield, an initial full-semblance pass is run. Intervals with reasonably good cement bond usually have both a peak for pipe arrival and one for formation arrival. If the cement bond is very good across the zones of interest, then the semblance peak associated with the formation will be dominant. Under these conditions, full semblance results will be more than adequate, and significant processing time can be saved.

In many situations, the pipe-semblance peak will be prominent enough to require use of a guided semblance. This requires tracking of the first waveform in the array. Using all eight receivers for semblance will yield a more robust answer. Care should be taken in partial bond situations, as some of the receivers closer to the transmitter may be affected by cement or pipe ring.

Comparison of full-semblance results and guided semblance will establish a level of confidence in the answer accuracy. In addition, this comparison will help adjust various settings for the final-semblance pass. This pass is run at a very high resolution. The resulting slowness trace is edited by tracking on the semblance peak as needed, Fig. 3.

Fig. 3. Results of final semblance pass are shown on Track 4.

COMPARING CASED-HOLE SONIC AND PND-S POROSITY

The combination of data from pulsed neutron and array sonic logs provides enough information for formation evaluation, seismic correlation and, in some instances, rock properties.

Because no openhole logs were run, PND-S pulsed-neutron logs were run primarily for the purpose of obtaining neutron- and density-type porosity data. Comparison of the sonic porosity with porosity values derived from the pulsed-neutron log exhibited the expected petrophysical relationship and, thereby, validated results from both logs, Fig. 4.

Fig. 4. Log display showing combination of neutron (green) and density (red) porosities derived from pulsed-neutron tool data in Track 3. Sonic (blue) porosity validates both results.

SHEAR SLOWNESS FROM REFRACTED MONOPOLE WAVEFORMS

The primary purpose in developing a slim monopole array tool was for acquisition of compressional slowness in through-tubing applications. Experience with the tool thus far has shown that, under the right conditions, shear slowness can also be acquired.

According to Snell's Law:

For the shear component of the refracted wave to enter the wellbore and be detected by the receivers, its slowness must be faster than slowness of the wellbore fluid. This relationship is the same in both openholes and cased holes.

In one of the wells studied, the wellbore intersected some formations where the shear component of the refracted waveform was discernible in the log presentation shown in Fig. 5. Because shear arrival in the waveform is farther away from the pipe ring than the compressional arrival, it can actually be easier to track. The shear component can be discernible even when cement bond degradation allows pipe arrival to interfere with compressional arrival.

Fig. 5. Log example of SMA cased-hole semblance processing. Track 1 contains gamma-ray (dashed green), Vp/Vs ratio (red), and Poisson's ratio (dotted blue). Track 2 contains compressional slowness and integrated travel time (blue) and shear slowness and integrated travel time (green). Track 3 presents a VDL display. Track 4 contains compressional slowness semblance processing, and Track 5, shear slowness semblance processing.

SMALL DIAMETER GAS WELL LOG EXAMPLE

The second case history represents an example of how a small-diameter tool can provide porosity information when borehole conditions make openhole logging difficult. Due to drilling problems in this deep and hot well, the openhole logging string was unable to pass below 20,750 ft, thus missing the important payzone below. The SMA tool, run after casing was set, registered a bottomhole temperature of 432°F. Although it was run primarily for cement bond and channel identification, Fig. 6a, careful processing produced a reasonable compressional slowness and porosity log, Fig. 6b.

Fig. 6. Through-casing CBL-5-ft VDL log presentation (a). Compressional slowness (blue) computed after semblance processing (b). The 3-ft VDL is in Track 3. Notice high pipe arrival content of the VDL.

The low amplitude of the compressional slowness semblance shown in Fig. 6b is due to the presence of the pipe arrival. Note that the 3-ft VDL in Fig. 6b has more pipe energy than its 5-ft VDL counterpart in Fig. 6a. Frequency filtering and manual tracking were necessary to render the compressional slowness.

OTHER RELATED STUDIES

In the SPWLA paper referenced in the Acknowledgment, log comparisons in small and large casings were made in a study from the EPA testing facility in Ada, Oklahoma. A test well with casing sizes of 4-1/2 -, 5-1/2 -, 7-, 8-5/8 - and 9-5/8 -in. was cemented with carefully designed, man-made and naturally occurring channels. During the late 1980s and early 1990s, service companies were invited to run their tools for identifying and characterizing these channels.

A 2-3/4 -in. Sector Bond tool was run in this well in 1992. Ten years later, the Slim Sector Bond was run, and results from the two tests were compared. In summary, the SSB can provide CBL and VDL logs in large casings (9-5/9 in.). Its 3-ft amplitude produces identical CBL to the larger-diameter Sector Bond tool, but it has lower circumferential channel resolution. The SSB offers an alternative logging option where the tubing cannot be removed from the well for mechanical or economical reasons.

The referenced paper also describes and illustrates how the stuck-pipe log complements the free-point indicator in fishing operations and provides important information to determine the amount of washpipe necessary. It is also helpful in estimating time needed for completing the fishing operation. An example from a West Texas well was used. The stuck-drill pipe log presentation is similar to typical CBL and VDL logs. The difference between the stuck-pipe and cement-bond logs is in the interpretation. On a CBL log, a typical value of 55 mV or 30% from the maximum on the amplitude is considered bad bond, while, on the stuck-pipe log (SPL), the same 30% amplitude value in free pipe is considered stuck pipe.

CONCLUSIONS

When openhole logging is undesirable, impractical or impossible due to borehole conditions or economics, cased-hole logging offers an alternative means for formation evaluation. A new through-casing, hostile environment monopole-array acoustic device can provide reliable formation compressional slowness in small-diameter wells with good cement bond.

Data processing ensures that only good-quality compressional and, in some instances, refracted shear data are delivered. Combining data from the SMA with pulsed neutron capture data allows correlation and accurate interpretation in gas-bearing formations. This slimhole tool provides, at the well site, high-quality cement-bond evaluation in small-diameter casing and can also provide good quality results in large-diameter casing, albeit with lower resolution. 

ACKNOWLEDGEMENT

The authors thank the management of BP and Precision Drilling for allowing publication of this article. Thanks to Stephen Prensky for editing of the original paper. Thanks to Emery Baca and Freddie Sellers for the data acquisition and to Marek Kozak for his help in the data analysis. Thanks also is expressed to the oil companies for their permission to publish their log examples. This article was prepared from information in the paper, “Through-tubing hostile-environment acoustic logging tool,” presented at the SPWLA 45th Annual Logging Symposium, Noordwijk, the Netherlands, June 6 – 9, 2004.

LITERATURE CITED

¹ Chudy, S., G. MacIntyre and P. Schuh, “Cased hole acoustic logging – A solution to a problem,” Transactions Paper I, SPWLA, 36th Annual Logging Symposium, Paris, June 1995.

² Tello, L. N., T. J. Blankinship and G. T. Alford, “Comparison of F-statistic, Incoherence and Semblance as methods to process digital array sonic logs,” Transactions Paper E, CWLS 13th Formation Evaluation Symposium, Calgary, Alberta, September 1991.

³ Tello, L. N., et al., “A dipole array sonic tool for vertical and deviated wells,” Paper 56790, SPE Annual Technical Conference and Exhibition, Houston, October 1999.

⁴ Precision Drilling Services, 2001, Chapters 1 and 5, Pipe Recovery Instructions Manual.

⁵ Tello, L. N., et al., “New efficient method for radial cement bond evaluation,” First Joint Symposium of the CSPG and CWLS, Calgary, Alberta, May 28 – 31, 1995.

THE AUTHORS
	Lucio N. Tello, a project manager for Computalog, Precision Drilling, received a BSc in electronics from the Universidad Distrital in Colombia and an MSc and DSc from the University of Texas in Arlington, both in applied physics. He has been designing sonic and ultrasonic tools for more than 25 years.
	Daniel L. Long, a senior log analyst for Computalog, Precision Drilling, earned a BSc in physics and a BSc in mathematics from Eastern Oregon State University in 1980. He has 23 years' experience in wireline logging and log analysis. Mr. Long is a member of SPWLA and SPE and was vice president of the Bakersfield SPWLA chapter in 1990.
	Thomas J. Blankinship, acoustic group leader for Computalog, Precision Drilling, Ft. Worth, Texas, graduated from Michigan Technological University in 1981, with an MSc degree in electrical engineering.
	Robert L. Horine has worked for BP for 15 years. He is a graduate from the University of Tulsa, Oklahoma and the University of Utah.