Drillstring sub cuts torque and casing wear

David William Rodman
Security DBS
Aberdeen

George Swietlik
Pilot Drilling Control
Lowestoft, England

This drillstring sub effectively reduced torque losses and helped improve torque transfer to the bit on some of the extended reach wells drilled on BP Exploration's Miller development in the U.K. North Sea.

Problems with excessive torque during drilling of the Miller field template wells led to a concurrent engineering design and testing program to develop a tool to minimize torque losses. When available technology was judged inadequate to solve these problems, a torque-reduction sub was designed, tested, and used in this series of critical, high-angle, deep, platform wells.

In extended reach drilling, current drilling equipment often reaches its service limits. Higher torque inputs from the top drive, larger angular twist and fluctuations on drill pipe, tool joint fatigue, and casing wear result from the frictional forces between the rotating pipe and the low side of the hole.

Earlier attempts at addressing the torque problems included modifications to bottom hole assembly (BHA) components and the use of tools such as roller reamers and nonrotating stabilizers/roller stabilizers. The results from these field trials were questionable and did not address the concerns on casing wear. They did not take into account that the BHA lengths were considerably less than the total length of the drillstring, and the summation of torque losses from frictional effects was greater in the interval over the drill pipe length.

Earlier mathematical analysis of torque profile clearly showed that the section of hole from the kick-off point to the end of build provided the greatest opportunity for reductions in torque. With this understanding, a new deviation plan was developed through the use of two build-up phases separated by an intermediate tangent section and also by lowering the kick-off point. These techniques were successful in easing the situation, but they did not address the underlying mechanical problem or casing wear. Other work on casing wear using different mud systems and aluminum tool joints also met limitations.

Several obstacles had to be overcome in the design and development of the solution. The service industry responded with several potential solutions; however, these failed to provide the minimum required objective of reducing off-bottom torque by 30%, combined with durability and reliability, to give a minimum of 200 rotating hr between service or laydown.

Given this situation and the impending requirement to drill the extended reach wells on BP Exploration's Miller development, a team (operator, service company, and design consultant) was formed to engineer the solution concurrently during the drilling of the Miller development wells.

Background

The Miller development team recognized from the template drilling experience that the naturally high field torque values and top drive power limit would restrict rate of penetration and well design and affect casing wear predictions. Initial investigations indicated that off-bottom torque losses accounted for a large percentage of deliverable torque. BP Exploration asked several service companies to try to solve these problems.

Adjusting oil/water ratios, putting additives in the mud, including motors in the BHA, adding roller reamers, using nonrotating stabilizers, and using various casing/drill pipe protectors were all tried with limited effect.

The majority of off-bottom torque was generated through the build-up section of hole. Therefore, an integral part of the drillstring in some form or another through this section would have the greatest effect on reducing the problem of torque loss, and thus casing wear.

Service companies were quick to come up with solutions that could perform part of the function, but the reliability and durability of these solutions were questionable.

Drill pipe bearing sub

The drill pipe bearing sub was developed to run as an integral part of the drillstring between tool joints and provide an ideal bearing surface at strategic points in the string. Initial tests indicated off-bottom torque reductions as high as 25%.

The major flaw in this design was that the mandrel was not a single piece (Fig. 1 [34009 bytes]). Thus, the number of joints in the string increased dramatically. Any bending movements or stresses would occur across the internal joint, the weakest joint in the string. A break in this joint could prove extremely difficult to fish. This was subsequently the demise of this solution.

Lo-Torq sub

The Lo-Torq sub was designed to reduce downhole torque. It can be positioned at any point in the drillstring without causing disruption to the BHA design criteria (Fig. 2 [28462 bytes]).

This type of hard banding (Armco) was subsequently used on the drill pipe tool joint itself as a means of reducing the number of tool joints in the string. This was only a minor solution.

By then changing the oil/water ratio to 80/20 and 90/10, casing wear was reduced significantly, but torque was still a problem. Even with the addition of Lubra-beads, the off-bottom torque still remained significant.

Nonrotating protector

The nonrotating drill pipe/casing protector (NDPP) consists of a rubber sleeve held in place axially by aluminum thrust collars. The collars are secured to the drill pipe by bolts instead of the more common wedge pins (Fig. 3 [37818 bytes]).

This type of protector has been one of the most commonly used and successful throughout the North Sea. It is also the tool against which the comparisons have been made.

The one major drawback of this tool is the potential for it to strip, leaving the rubber element downhole with consequential fishing or directional drilling problems. It is also necessary for an engineer to be on site to check the tools when they are tripped in or out of the hole. Time spent on placing the NDPPs in the string is also considerable.

This tool significantly reduced torque (around 30%), and casing wear was also reduced significantly.

As this was the most effective tool on the market in the early 1990s, operators tolerated the time and lost tool problems in order to achieve their extended reach drilling objectives. NDPPs were used on the Miller development from Well A6 to Well A11, until the introduction of the DSTR sub.

Low-torque drill pipe

Low-torque drill pipe consists of a single section of pipe with three integral stabilizer blade-like structures (Fig. 4 [29450 bytes]). The entire section is milled/turned from a single piece of steel bar.

The low torque ability of the tool is most prevalent when it is initially run in the hole. The reduction in torque at surface is derived from the stimulation of the cuttings beds just below the build section. Once these beds are removed the torque remains constant, but because of the contact area of the three blades (even though they are coated with a new type of hard-banding), significant torque levels can still be produced.

This tool was excellent for stimulating the cuttings beds, but its use was definitely restricted to cased hole. In open hole, it is possible that key seating may pose a problem.

The low-torque drill pipe was introduced around the same time as the DSTR, sub and the two are proving to be very effective when used in combination.

DSTR sub

The initial design allowed for the drill pipe tool joints to be lifted off the side of the casing and rotated through a static, rigid bearing shell. Because numerous tools would be required to appraise their operational effectiveness and service potential, the industry was canvassed to assess the opportunities and financial risks for such a project.

As these tools would be interspersed in the drillstring and form an integral part of the rotating string mass, the design called for a strong, single-piece mandrel with torsional and tensile strengths greater than those of the comparable pipe. Thus, they would not be the weak link.

In addition, the high bending forces transferred through to the mandrel body called for the design to accommodate cyclical bending and stress concentrations in the bearing area. This is now subject of a separate patent.

The objective was to considerably lower the frictional forces by incorporating rolling bearing contact faces instead of plain bearing surfaces (Table 1 [7059 bytes]). These minimum values exist under ideal operating conditions. They can, however, be exceeded as an operation proceeds in unfavorable conditions, as demonstrated by the tests at the research center. Inadequate lubrication or loss of lubrication under load will accelerate this process.

For all bearings operating under conditions of mixed or dry friction, however, there will be slight differences between contact and sliding friction, and there will be no absolute freedom from stick-slip behavior. Under normal conditions, with good uncontaminated surfaces, this can be neglected.

Based on these design considerations, a tool was built and tested against the NDPP with a section of Armco-coated tool joint as the base line. These tests were carried out at BP Exploration's test facility in Sunbury, U.K.

The DSTR subs were initially introduced for use on the Miller development in Well A11. Ten tools were manufactured and tested for durability. The original design of the tool consisted of only two ball bearing raceways.

This design was unable to withstand the required 200 rotating hr expected before being laid out, so a second design was introduced which included extra roller bearing raceways. This design was most effective because the roller bearings were expected to take about 70% of the radial forces and the ball bearings would take the remaining radial forces plus any axial forces.

While this second design did achieve the required 200 hr between servicing, the primary components were still undergoing significant wear. To overcome this, a program of concurrent engineering was started to improve the durability and serviceability of the tools while ensuring availability of tools for extended reach drilling on the Miller project.

In effect, these improvements were made during the service time between wells and consisted of the following:

• Modification of the mandrel and sleeve design to include additional roller race locations for use when original locations experienced wear

• Use of replaceable races in roller bearing locations in both the mandrel and sleeve

• Inclusion of a simple barrier seal mechanism to obstruct the flow of drill cuttings into the bearing areas.

These improvements were engineered and implemented in the original tools during the period of the Miller development drilling, resulting in the current tool design (Fig. 5 [29997 bytes]).

DSTR sub spacing

Well-simulation programs helped determine that the majority of side load (thus torque loss) occurred over the build section (Fig. 6 [29807 bytes]). Therefore, placing one sub every second or third tool joint to cover the build section for the duration of the bit run significantly reduced torque levels.

The dogleg severity is the most important factor in determining the spacing of the subs. The objective of the subs is to lift the tool joint off the casing wall so the drillstring rotates freely within the well bore with a minimum effective coefficient of friction over the highest side loading interval (Fig. 7 [33486 bytes]).

In most cases where excessive side wall loadings are shown on the well simulation, the dogleg exceeds 2°/100 ft. Spacing of 60 ft would be advised throughout the build section to attain torque reductions around 40%.

In cases where the dogleg is less than 2°/100 ft with a tangent angle of less than 60° and total drilled depth less than 19,000 ft, the use of the tool in some cases becomes marginal. Nonetheless, there are vertical hole applications in areas such as the western desert and the Rockies where there is an attrition on tool joint connections because of the ragged hole conditions which could be minimized or eliminated by the use of these tools.

The initial running parameters were set at a minimum of 200 rotating hr per tool before being laid out and replaced.

With the latest improvements in the raceway design, the subs are capable of running for at least 1.44 million revolutions (120 rpm for 200 hr). Table 2 [13736 bytes] illustrates the mechanical properties of the three sizes of sub available so that comparisons with drill pipe can be made.

Laboratory tests

The tool joint protector test facility is based on the flexible pipe test equipment at Sunbury Research and Engineering Centre.

The equipment was adapted so that drillstring tool joint protectors could be tested under simulated operational conditions with the tool joint traveling around a bend and, therefore, being subjected to a side load.

The test facility enabled sections of drill pipe/tool joint to be continuously rotated inside a length of casing which was restrained to one degree of freedom normal to the center line in the horizontal plane. A side load was then applied, via an hydraulic servo-controlled actuator and load cell, halfway along the pipe/casing. Throughout the testing process, oil-based drilling mud was circulated through the annulus between the casing and the pipe.

The test facility specifications were as follows: 0-30 kN side load, 15 kw motor power, 1 kNm torque, 123 rpm pipe speed, 5-in. or 65/8-in. pipe diameter, 2,844-m pipe length, 103/4-in. casing diameter, 1,750-m casing length, and 60 l./min oil circulation.

The pipe speed, casing temperature, side load, and power to drive the motor were logged every 30 sec. The torque to drive the DSTR sub was estimated as follows:

 T = (30µ/pi) 3 (P/S - Po/So)

In this equation, T is the torque to drive the DSTR sub in Nm, P is the power to drive the motor in W, h is the combined efficiency of the motor and drive train (estimated as 0.85), S is the drill pipe speed in rpm, and the subscript o denotes zero side load conditions

As a base line comparison for the DSTR sub and the NDPP, a hard-banded 65/8-in. drill pipe tool joint was run under normal test conditions from 0 up to 3,000 lbf side load. The test was run for approximately 1 hr at each side load. These results were then used as a base line.

Fig. 8 [20076 bytes] shows the variation in torque against side load for the DSTR sub, the NDPP, and the hard-banded tool joint. All the torque values on this plot are averages from the endurance tests after the torque had settled to a constant value. The coefficient of friction (i.e., the conventional first order coefficient, µ = F/N) for the tool joint varied. The maximum coefficient measured was 0.13 at 3,000 lb_f. (NB - µ for lubricated steel on steel is 0.16.) Table 3 [5536 bytes] shows the industry-accepted values for m in drilling wells.

This variation of coefficient of friction with side load resulted from the variation in the contact and lubrication conditions. The tool joint, when running on the casing at lower loads, acts like a lubricated friction bearing. At high loads, more of the lubricant (oil-based mud) is excluded from between the tool joint and the casing. Some parts of the joint may run metal on metal, giving rise to a higher coefficient.

Lab results

During the test it was not possible to rotate the hard-banded drill pipe above 3,000 lbf side load. Severe vibrations were present from 2,000 lbf side load, and damage had already been inflicted on tool joints on a previous test.

Because of the estimated drive train efficiency, absolute values of torque are approximate, but comparisons of relative values are reliable for the DSTR and NDPP.

The DSTR sub was assembled between the tool joints of the 65/8-in. pipe such that the center of the sub was positioned central in the test equipment and at the point of the applied load.

Up to a side load of 1,300 lbf the required torque to drive the tool was approximately zero. Under higher loads, it was realized that the outer rigid steel sleeve remained reasonably static inside the casing, negating the need to investigate other materials to counter casing wear.

With the NDPP tool, the flexibility caused the rubber bearing interface with the pipe to deform. This loss of free running would cause the pipe to start vibrating from a side load of 2,500 lbf.

The variation in the torque with time for a given side load was found to be negligible.

The maximum temperature rise during the tests was approximately 16° C.

Miller trials

After complete analysis of the test results from Sunbury and the durability test on Well A11, 100 DSTR subs were manufactured and sent out to be field tested.

The profiles of each well in the development were very similar (Fig. 9 [20993 bytes]), with the build ranging from approximately 40° for production wells up to approximately 60° for injection wells.

• Miller A12. Fifty subs were run at 60-ft intervals through the build section in Well A12. Although torque was reduced significantly, it was argued that these results were due to formation and bit adjustments.

  It was only on well A13 where off-bottom torque readings alone were taken to show the torque losses occurring solely through the string.

• Miller A13. The results for Well A13 indicate off-bottom torque readings while tripping only (Fig. 10 [26647 bytes]). As can be seen from the first two trips in the 121/4-in. section where no subs were used, the torque steadily increased with depth. On the third trip, as subs appeared across the build section, the off-bottom torque remained constant as depth increased. Torque reductions in excess of 40% at total depth were recorded. The subs were thereafter used as standard drillstring equipment on the Miller extended-reach wells.

• Miller A16. Although BP was completely satisfied with the results achieved from the tools, the company requested a test to determine the optimum number of subs required to reduce the torque to an acceptable level.

The mean figure for torque at 15,000 ft was 23,000 ft-lb with no subs; with 26 subs through the build the reduction in off-bottom torque was 18% (Fig. 11 [23729 bytes]). With 50 subs in the string, off-bottom torque was reduced by 24%. With 70 subs in the string, torque was reduced by 36%.

For all three runs, the tools were spaced at 60-ft intervals, indicating 70-100 tools would be required for subsequent wells. No field-test results were generated for wells A17-A21, although 70-100 tools were used on each well.

Casing wear results

The main emphasis of the DSTR sub development for the Miller project was to reduce drillstring torque. Casing wear reduction was a secondary benefit of the DSTR sub.

Casing wear measurements were obtained from another BP project in the Leven field, Block 30/17-A26. Although the well profile was similar to those at Miller, the prime concern was high casing wear. Fifty subs were used on a 90-ft spacing over the 2°/100 ft build section.

Of significance was the reduction of the recovered casing steel particles when the subs were run (Fig. 12 [23507 bytes]). At 120 rpm drilling speed, steel recovered remained in the low area whereas on previous wells, recovered steel ranged between the medium and high wear factors.

Future developments

The DSTR drag tool is presently under development. The new design consists of the basic DSTR mandrel and outer sleeve assembly, together with the addition of bearings on the outer sleeve. This design is expected to reduce drag by similar margins.

A cuttings bed impeller is an adaptation of the single piece DSTR sub mandrel with the addition of a modified, rotating sleeve with a blade configuration. This design allows for stimulation of the cuttings beds which normally settle just below the build section. By stimulation of these cuttings beds, possible twist-off and other associated problems can be reduced.

Recommendations

To meet the growing demands on extended-reach well design for longer stepout capability, it has become essential to develop tools within the drillstring which minimize torque losses and ensure that torque transferred to the bit is kept as high as possible.

The existing range of DSTR subs covers 5-in., 51/2-in., and 65/8-in. drill pipe sizes, which are most commonly used in extended reach wells today.

The use of DSTR subs extends the limits of a well design while maintaining sufficient power to the bit. It is no longer necessary to over-engineer the casing to withstand high wear factors, nor is there a requirement for excess capacity top drive units to compensate for frictional torque losses in the drillstring.

Further investigation is required to confirm reductions in drillstring torque fluctuation, drillstring dynamics, and slip-stick conditions when using DSTR subs in highly deviated wells.

The DSTR subs are designed such that all mechanical properties are to American Petroleum Institute specifications with limits greater than those of the string in which they are connected.

Acknowledgment

The authors thank BP Research and BP Exploration for sponsoring this project and many others for their valued contributions towards the compilation and preparation of this article.

Bibliography

1. Kwakwa, K.A., Batchelor, A.S., and Clark, R.C., "Assessment of the Mechanical Stability of High Angle Wells in Block 22/11, Nelson Field Discovery," Society of Petroleum Engineers Drilling Engineering, March 1991.

2. Johancsik, C.A., Friesen, D.B., and Dawson, R., "Torque and Drag in Directional Wells-Prediction and Measurement," Journal of Petroleum Technology, June 1984.

3. Brett, J.F., Beckett, A.D., Holt, C.A., and Smith, D.L., "Uses and Limitations of Drillstring Tension and Torque Models for Monitoring Hole Conditions," SPE Drilling Engineering, September 1989.

4. Jenkins, R.W. and Patricks, A.N., "Drilling The Deepest Horizontal Well in the Gulf of Mexico, South Marsh Island 239 D-10," Paper 23879, presented at the International Association of Drilling Contractors/Society of Petroleum Engineers Annual Drilling Conference, New Orleans, February 1992.

5. Banks, S.M., Hogg, T.W., and Thorogood, J., "Increasing Extended-Reach Capabilities Through Wellbore Profile Optimization," paper presented at the IADC/SPE Annual Drilling Conference, New Orleans, February 1992.

6. Tollefsen, S., Graude, E., and Svinddal, S., "The Gullfaks Field Development Challenges and Perspectives," Paper 25054, presented at the SPE European Petroleum Conference, Cannes, France, Nov. 16-18, 1992.

7. Larsen, T., "A Study of the Critical Fluid Velocity in Cuttings Transport for Inclined Wellbores," master's thesis, University of Tulsa, 1990.

8. Pilehvari, A., "Modeling of Cuttings Transport in Highly Inclined Wellbores," University of Tulsa advisory board meeting, May 1992.

9. Iyoho, A.W., "Drilled Cuttings Transport by Non-Newtonian Drilling Fluids Through Inclined, Eccentric Annuli," doctoral thesis, University of Tulsa, 1980.

10. Sifferman, T.R., and Becker, T.E., "Hole Cleaning in Full-Scale Inclined Wellbores," Paper 20422, presented at the SPE Annual Technical Conference and Exhibition, September 1990.

The Authors