Case histories show real-time information reduces uncertainty

Drill-bit seismic technology-Conclusion

Richard Meehan
Sclumberger Cambridge Research
Cambridge
Les Nutt
Schlumberger Wireline & Testing
Houston
Neder Dutta
BP Exploration
Houston
James Menzies
Lasmo plc
London

Drill-bit seismic technology provides useful real-time information that reduces uncertainty for the driller.

This second part of a two-part series provides real examples, describes the current state of technology and experience (see box), and discusses techniques that will add value through the incorporation of additional information sources.

To date, this technology has been used on more than 50 wells in all areas of the world, including a deepwater job in 1,200 m of water. As of March 1998, there were five ongoing jobs in Oman, Indonesia, Malaysia, Gulf of Mexico, and South America.

The following examples demonstrate the application of the drill-bit seismic technique and show benefits that can be brought to the drilling process.

Well A

Well A was drilled by BP Exploration and Statoil AP offshore Vietnam in 1996.¹

This exploration area is characterized by normally pressured clastics in the shallow sections.

These are followed by under-compacted shales that are often overpressured. The pressure ramps in this area can be very severe, sometimes as high as 3 psi/ft. The overpressured shales are mechanically weak, thus, the safe mud-weight window for drilling this section is quite narrow.

Depth predictions, based on surface seismic velocities, carry uncertainties of up to 300 m. Other wells drilled in the area had significant problems, some failing to meet their targets.

The most critical factor in drilling these wells is to set the 133/8-in. casing precisely at the top of the pressure ramp. Without reliable real-time estimates of the pore pressure and depth to formation tops, the well plan would have required a contingent 16-in. casing string and associated underreaming of the open hole section, at a cost of $3 million.

Based on previous experience of the seismic-while-drilling technique, the operators were confident that the real-time information offered by drill-bit seismic would render the 16-in. liner unnecessary. Eliminating it from the well plan saved $500,000.

drill-bit seismic data were acquired from a depth of 1,100 m down to 3,550 m. The onset of overpressure is marked by a major change in lithology and fossil content, and a recognizable seismic marker.

The drill-bit seismic data were integrated with measurement-while-drilling (MWD) data, drilling parameter data, cuttings, and biostratigraphy analysis. The average velocities derived from the drill-bit seismic data were different from those estimated from the surface seismic. These are shown in Fig. 1 [40,785 bytes].

Also shown are the average velocities derived from the conventional wire line borehole-seismic survey run after the well reached TD.

The interval velocities, or slowness, can also be calculated from the drill-bit seismic data. These are shown in Fig. 2 [44,084 bytes] along with the slowness from the wire line sonic log acquired after drilling.

The onset of overpressure occurs at a depth of about 2,500 m. Confidence in the look-ahead images derived from the data supported the prognosis of a clean mudstone, positioned below the seismic marker at the top of the overpressure zone, and allowed the 133/8-in. casing point to be pushed deeper into the zone.

In the deeper section of the well, the drill-bit seismic data, in conjunction with other information, confirmed that it would not be possible to reach the 95/8-in. casing point planned for the bottom of the overpressure ramp without setting the contingent 113/4-in. liner.

The real-time nature of the technique ensured this became evident before the first planned coring point was reached. This allowed the 113/4-in. liner to be set deep enough, ensuring the 95/8-in. casing could be set as planned, but close enough to the desired coring point to reduce the amount of open hole section during the coring operations.

Well B

Well B was drilled in July 1995 by Lasmo plc and its partners, Union Texas Petroleum and Itochi Oil Exploration Co. Ltd. The well was also located offshore Vietnam, at a water depth of 125 m.

In this case, an array of 24 hydrophones was used. drill-bit seismic data were acquired from 1,500 m to 3,750 m, and data quality was very good throughout. Fig. 3 [124,345 bytes] shows the final stacked one-way time wavefield for the 121/4-in. section.

The direct arrival is strong, and since only hydrophones were used, the sea-surface multiples are clearly visible. Fig. 4 [149,652 bytes] shows part of the surface seismic section. The inset section is a corridor stack calculated from the drill-bit seismic look-ahead image. The correlation with the surface seismic is very good.

The real-time information supplied by the drill-bit seismic technique, when combined with other measurements such as drilling parameter and cuttings analysis, allowed the 17-in. hole to be drilled to just over 2,900 m, leading to the longest 133/8-in. casing run in this province.

The 121/4-in. section was pushed down to 3,750 m, allowing TD to be reached in the 81/2-in. hole, eliminating a planned casing string at a saving of almost $2 million. This was the hottest well drilled in the area, with a bottom hole temperature greater than 400° F.

Well C

Well C was drilled by Norsk Hydro AS in 384 m of water, in the Norwegian sector of the North Sea early 1997.² The well was in an area where the surface seismic data quality was poor, primarily because of the presence of gas-rich shales (gas chimney effect).

The excess gas caused attenuation of the seismic energy, making it difficult to estimate the time/depth relationship. This in turn led to large uncertainties in the predicted formation tops and pore-pressure estimates. The prognosed pore pressures were sufficiently high to require the inclusion of a contingent 16-in. casing string in the well plan.

Because of the uncertainties, it was decided to use the drill-bit seismic technique in conjunction with other methods, in particular a basin modeling and inversion technique, and drilling parameter analysis.²

Fig. 5 [99,532 bytes] shows the final stacked one-way time, drill-bit seismic wavefield for part of the 171/2-in. section of the well. The direct arrival from the bit can be clearly seen. Picking the direct arrival provides the time-to-depth relationship.

Differentiating this with respect to depth provided an estimate of the local formation velocity. The data were also processed to produce look-ahead images, indicating upcoming formation tops, and providing a tie with the surface seismic.

The formation velocity estimates were used, along with drilling parameter analysis, to determine the pore pressure at the bit. In addition, the formation velocity estimates and look-ahead images were used in conjunction with the basin modeling to predict pore pressure ahead of the bit.

Although abnormally high gas values (greater than 20%) were encountered, the pore pressure as indicated by the drill-bit seismic data and the other techniques allowed drilling to continue past the contingent 16-in. and the planned 133/8-in. casing points. High confidence in the predicted pore pressure allowed optimum mud weights to be used.

The final depth of the 171/2-in. section was determined by the amount of 133/8-in. casing on the rig, and was some 500 m deeper than originally planned. This led to a saving of 6 days rig time and a 16-in. liner string.

Drilling-hazard prediction

The above examples demonstrate how real-time information provided by drill-bit seismic can be used with drilling parameter data, cuttings analysis, modeling techniques, and the surface seismic data to greatly improve the overall efficiency of the drilling process.

The drill-bit seismic technique can also be used in conjunction with conventional borehole seismic surveys. The combination of these two methods allows real-time prediction of depth to interpreted overpressured zones, and can even provide a recommendation for minimum mud weights.³

One of the most popular and important applications of traditional borehole-seismic, vertical seismic profile (VSP) data can be used to predict overpressure ahead of an intermediate TD.

This is achieved by inverting the VSP data for acoustic impedance. VSP data are more suitable for this task than surface seismic data since they usually have higher bandwidth and a better signal-to-noise ratio.

A seismic trace is an indication of variations in acoustic impedance. These variations depend upon formation velocity and density. A decrease in acoustic impedance can indicate an increase in porosity, thus, it indicates a potentially overpressured zone.

There are various techniques for inverting VSP data for acoustic impedance.⁴⁵ The output is generally in the form of a prediction of acoustic impedance vs. two-way time. For drilling, however, it would be much more useful to plot acoustic impedance against depth ahead of the bit.

Using an empirical velocity-density relationship such as that proposed by Gardner,⁶ calibrated by data from nearby wells if available, allows the acoustic-impedance profile to be converted to interval velocity. It then becomes fairly straightforward to convert the two-way time values to depth values.

The formation velocity prediction can be transformed to a pore-pressure estimate, or minimum mud-weight recommendation, by using local information on the relationship between formation velocities (or slowness) and pore pressure. If there are insufficient local data, various empirical methods have been suggested in the literature, for example the well known Hottman-Johnson relationship.⁷

The accuracy of the predicted depth to the interpreted overpressure zone depends upon the validity of the assumptions in the velocity-density relationship and the efficacy of the VSP inversion technique.

However, as drilling progresses from intermediate TD, the point at which the VSP data were acquired, the real-time time-to-depth information from the drill-bit seismic technique can be used to update the depth-to-hazard prediction.

The top part of Fig. 6 [47,828 bytes] shows the acoustic impedance inversion calculated from an intermediate wire line VSP. The sudden drop at about 2.2 sec two-way time is the top of the overpressured zone. The bottom part of Fig. 6 shows that the initial depth estimate for this hazard is 2,707 m.

As more time-to-depth information becomes available, this prediction is updated. The dashed line in Fig. 6 shows an updated depth prediction; the depth to the hazard is now 2,753 m. The closer the bit approaches the hazard, the more accurate the prediction becomes.

The technique described above depends upon a successful inversion of the wire line VSP data. It would be much more convenient if the drill-bit seismic look-ahead image could be inverted for acoustic impedance, thus, eliminating the need for the intermediate wire line VSP along with its associated rig-time requirement.

However, the poorer signal-to-noise ratio of drill-bit seismic data, and the lack of control over the source signature mean the data are not usually suitable for inversion. As the methodologies for acquiring and processing drill-bit seismic data improve and evolve, this situation will improve.

Looking ahead

The drill-bit seismic method is still evolving. Techniques for improving data acquisition and processing are being developed. The operational limits for water depth and well deviation are being pushed further and further outward.

Perhaps the biggest challenge will involve improving the integration of drill-bit seismic data with other measurements and methods for optimizing drilling performance. One obvious example would be to combine the real time check-shot output of drill-bit seismic with sonic logging-while-drilling data. This would enable the generation of real-time calibrated sonic logs and synthetic seismograms for correlation with surface seismic data.

Bringing together information from many different sources can optimize the drilling process. To fully realize the potential benefits, the professionals involved in well construction must work together effectively.

In the examples above, an essential component for success involved using a properly integrated team of geologists, geophysicists, and drilling personnel. If drill-bit seismic can encourage real cooperation between disciplines, it will have played another part in drilling optimization.

References

Jackson, M., and Einchomb, C., "Seismic while drilling: Operational experiences in Viet Nam," World Oil, Vol. 218, No. 3, March 1997, pp. 50-53.
Doyle, E.F., "Recent experiences in Seismic-While-Drilling and new formation pressure techniques on a Norwegian HP/HT well," presented at the 11th Annual Offshore Drilling Technology Forum, Aberdeen, Nov. 25-27, 1997.
Borland, W.H., Hayashida, N., Kusaka, H., Leaney, W.S., and Nakanishi, S., "Drill-bit seismic, Vertical Seismic Profiling, and Seismic Depth Imaging to aid drilling decisions in the Tho Tinh Structure, Nam Con Son Basin-Vietnam," presented at the Japanese Society of Exploration Geophysicists, Kyoto, Oct. 21-23, 1996.
Carron, D., "Well guided stratigraphic inversion of borehole and surface seismic sections," presented at the 58th International Meeting of the SEG, 1988.
Walker, C., and Ulrych, T.J., "Autoregressive recovery of the acoustic impedance," Geophysics, Vol. 48, No. 10, 1983, pp. 1338-50.
Gardner, G.H.F., Gardner, L.W., and Gregory, A.R., "Formation velocity and density-the diagnostic basics for stratigraphic traps," Geophysics, Vol. 50, No. 11, 1985, pp. 2085-95.
Hottman, C.E., and Johnson, R.K., "Estimation of Formation Pressures from Log-Derived Shale Properties," JPT, Vol. 17, June 1965, pp 717-22.

Current state of drill-bit seismic technology and experience

THE COMPONENTS AND CONDITIONS DESCRIBED below indicate the current state of experience in relation to drill-bit seismic techniques. However, it must be kept in mind that these components and processes do not necessarily impose fundamental limitations on the technology, and that improvements will push back the current operating envelope.

Bit type: In its current form, drill-bit seismic works reliably only when drilling with a roller-cone bit. This is because the drilling action of a roller-cone bit generates the required axial vibrations in the drillstring and P-waves in the formation.

Conversely, polycrystalline diamond compact (PDC) bits drill by shearing the rock, generating very little in the way of drill pipe axial vibrations. Thus, it is not usually possible to detect the bit-generated signal using an accelerometer mounted on the swivel or top drive.
Occasionally, measurable signal levels have been detected when using PDC bits, particularly when drilling hard rock, or when some of the PDC cutters are missing or damaged.

Water depth: The deepest commercial drill-bit seismic job so far occurred in 384 ft of water (November 1997), in the Norwegian sector of the North Sea.1 Recent tests in the Gulf of Mexico, however, have demonstrated that the drill-bit seismic technique will work in much deeper water. Good-quality data were successfully collected in 1,200 m of water, and deeper tests are planned.
Well deviation: The largest well deviation in a commercial drill-bit seismic job is currently 65°. This was recorded on land in Tunisia where the objective was to guide the well trajectory to the assigned target.2 In general, the limiting factor is the attenuation of the drillstring signal, caused by the interaction of the drillpipe with the well bore. The amount of attenuation depends upon the length and profile of the well. It should also be noted that a roller-cone bit is a dipole source, aligned along the axis of the drillstring. In a horizontal well, almost no compressional wave energy travels directly upwards to the surface.
Downhole motors: Good-quality data have been collected when drilling with downhole motors if the drillstring is also being rotated from surface. In sliding mode, signal quality is usually poorer because the drillstring signal is attenuated. Usable data can be acquired in most lithologies. Data quality is usually poor in very unconsolidated formations, particularly when drilling is by controlled penetration rates with low bit weight (less than 10,000 lb).
Lithology types: Usable data can be acquired in most lithologies. Data quality is generally poor in very unconsolidated formations, particularly when drilling occurs through controlled penetration rates using low bit weight on bit (less than 5 tons).

References

Doyle, E.F., "Recent experiences in Seismic-While-Drilling and new formation pressure techniques on a Norwegian HP/HT well," presented at the 11th Annual Offshore Drilling Technology Forum, Aberdeen, Nov. 25-27, 1997.
Van Derck, R., Beck, B., Belaud, D., and Underhill, W., "Drill-bit seismic, A service for well trajectory steering," presented at the Offshore Mediterranean Conference, Ravenna, Italy, Mar. 19-21, 1997.