IMPROVING OPERATIONAL PROCEDURES CAN REDUCE FATIGUE FAILURE IN 8-RD CASING

Michael L. Payne
ARCO British Ltd.
Poole, U.K.

Richard E. Leturno
ARCO Indonesia Inc.
Jakarta

Chris A. Harder
Vastar Resources Inc.
Houston

Understanding the factors that contribute to casing fatigue can improve well planning, material procurement, and field operations to reduce the likelihood of fatigue failures.

Many operators have experienced fatigue failure of 8-round (8-rd) drilling casing. This failure mode is unusual, unexpected, and difficult to prepare for in design.

This fatigue failure is important because of the cost of casing failures. Fatigue failures most often occur in drilling casings near the surface, thereby exposing operations to severe well control risks.

Key considerations in these fatigue failures include casing and cementing design, landing, drillstring/casing interactions, dimensions and gauging of 8-rd connections, casing ovality, and mill and field makeup control.

This type of failure is not new and has been described in the technical literature. Familiarity with this type of failure is currently low, however, leading to the need to reemphasize the problem and reinforce proper design and field practices.

THREADED CONNECTIONS

In drilling and production operations, tubular fatigue technology has almost exclusively focused on downhole drill stem and offshore marine components.

Drillstring component fatigue has been heavily researched, and work continues in the area. Advances include optimization of threadform geometry to reduce stress concentrations, cold rolling of thread grooves, the use of stress-relief grooves, and the enhancement of upset geometry (because of its correlation to drill pipe fatigue life).

Likewise, offshore tubular applications, such as drilling and production risers, pipelines, and tendons, have been the focus of intense engineering analysis and optimization. Advances here include threadforms with variable pitch and variable thread height to minimize stress concentrations and more sophistication in the manner by which stochastic variables (such as environmental loads and mechanical properties) are accounted for in design.

Unlike drill stem and o shore components, drilling casing is considered a static well bore component. Although casing has not been studied in great detail with regard to fatigue endurance, some pertinent research is available.

In 1981, researchers from Vallourec published a study on the fatigue endurance of a threaded connection for casing used for offshore pipelines.1 The work was performed on 5-in., 15-lb/ft N-80 casing with a premium connection frequently used for downhole casing and tubing. This study demonstrated that fatigue endurance of a threaded connection can be increased by increasing the wall thickness near the threads, gluing,the connection to increase the load transfer, and protecting critical fatigue areas from corrosion.

In 1982, Brown and Bartle of British Gas plc analyzed the fatigue failure of conductor casings on exploration wells in the English Channel. 2 Tidal currents up to 5 knots were causing vortex-induced vibration of the conductors, resulting in a series of failures. The work included evaluation of special connections and vortex spoiler mechanisms.

In 1988, Bruno reported on a series of fatigue failures caused by transit damage of tubulars." This study covered primarily line pipe failures and provided recommendations for proper transport of pipe to restrict motion and, hence, fatigue damage during transport.

In 1992, Sumitomo verified the likelihood of fatigue at the last engaged thread (7-in., 32-lb/ft casing with a premium connection) if precautions were not taken to enhance fatigue endurance at that point.4

In 1955, Texter presented a classic overview of tubular failures and their cause in the most specific publication on fatigue failure of 8-rd casing. 5 Texter analyzed case studies of "last engaged thread failures" and made several observations:

• All failures occurred above the top of cement.

• Straight holes produced more failures than deviated wells, which dampen drillstring/casing interaction. 0 The failures involved various grades of pipe and both seamless and ERW (electric-resistance weld) types.

The recommendations to eliminate this failure mode included the following:

• Using drill pipe rubbers to reduce drillstring/casing interaction

• Using round-Y threads (known now as 8-rd) instead of the earlier threadform design known as sharp-Y threads.

CASING FAILURES

Table 1 summarizes the following casing failure examples. These failures support the original observations by Texter concerning failure in straight holes above the top of cement.

However, 8-rd obviously does not provide a cure for this problem, despite its improvements relative to its predecessor, the now obsolete sharp-Y thread.

ONSHORE SOUTH TEXAS

A string of 13 1/8-in. surface casing was run to 2,000 ft. A 12 1/4-in. hole was then drilled to about 8,800 ft, and 9 5/8-in., 47-lb/ft S-95 long thread and coupling (LTC) protective casing was run in the well, which was essentially vertical. The bottom 1,500 ft of the string was cemented, and the casing was hung with 380,000 lb on the slips,

An 8 1/2-in. hole was then drilled to 10,700 ft, with the mud weight increasing from 10.8 to 14.9 ppg. Following a bit trip, the drillstring could not pass a tight spot at 536 ft; the casing had parted there. The upper section of casing above the part was retrieved, and the joint below the failure was backed out.

Inspection of the failed connection revealed a fatigue crack around the pin at the last engaged thread. The crack extended 9-in. around the circumference.

The 12 joints of casing above the failure were inspected. Mechanical properties of the pipe and couplings were found adequate. Three of the 12 joints, however, had cracks in the pins at the last engaged thread.

A casing caliper log was run. Substantial wear was found in the casing above the top of cement, raising concern about helical buckling. Engineering calculations verified that the string had buckled from a combination of inadequate cement sheath and landing tension.

Because the failed pin and all the cracked pins occurred on mill-end connections, the mill-end makeup procedures were reviewed. Significant inadequacies were found in the quality control of the mill-end makeup, and recommendations were provided for future orders.

Drilling through the buckled casing caused a large amount of interaction between the drillstring and casing. The resultant dynamic motions of the casing led to the fatigue cracks. The mill-end connections, rather than the field-end connections, had fatigued because of the higher torque used on them during makeup. All cracks were near the surface because of maximum tension loads.

Remedial recommendations involved proper landing procedures and mill-end makeup control.

OFFSHORE TEXAS

This well was drilled as a straight hole with a jack up rig in 81 ft of water. A string of 9 5/8-in., 36-lb/ft J-55 LTC casing was run and cemented in a 12 1/4-in. hole at about 3,500 ft. The casing was landed using a landing ring, so tension above string weight could not be applied.

After 18 1/2 hr of rotating time, the well began losing mud through the 30-in. x 13 3/8-in. annulus. The 9 5/8-in. casing was picked up and found free. The top four joints were pulled. The last joint retrieved had a coupling "looking" down, meaning that a mill-end pin looking up remained in the well.

The casing was backed off at 620 ft, and additional joints were removed. New casing was tied back into the well, and slips were set with 120,000 lb tension. A top cement job was then performed. The well was then completed in about 60 rotating hr.

During the abandonment procedures, the connections in the upper seven joints were found to have fatigue cracks at the last engaged thread. Four of these joints had cracks in the mill-end and three in the field-end.

Subsequent analysis estimated that the top of cement around the 9/8 in. was in the range of 1,000-1,600 ft. With this uncemented interval and no additional tension applied to the string, it was determined that this string had been subjected to helical buckling. The helical buckling, uncemented interval, light casing weight, and poor fatigue resistance of 8-rd casing explain the first fatigue failure. The subsequent series of failures in the tie-back string is not fully understood and raises questions about excessive drillstring/casing interaction.

COLORADO

A 17 1/2-in. surface hole was drilled to 515 ft, and 13 3/8-in. casing was run and cemented. The casing was drilled out with a 12 1/4-in. bit. The intermediate hole was drilled to 4,353 ft, and 9 5/8-in., 36-lb/ft J-33 short thread and coupling (STC) casing was run with a stage tool at 3,266 ft.

Because a caliper log was not run, the cement volume pumped was based on 100% excess of gauge hole. The cement job had full returns, and most of the spacer was recovered, implying cement was circulated close to surface. A cement bond log indicated the top of cement was at 730 ft for the second stage and 3,480 ft for the first stage. The casing was tested to 1,500 psi.

After the shoe was drilled out, the hole was unloaded with nitrogen and preparations made to drill with gas. When the 83/4-in. hole reached 7,600 ft, the surface samples became damp. The initial conclusion was that a formation was wet. Drilling continued to 7,810 ft, and the hole started becoming wet, indicating a possible casing failure.

The drillstring was picked up nine joints and then stuck at 5,050 ft. The drill pipe was backed off at 5,016 ft, leaving a 2,500-ft fish in the hole. During jarring operations, the fishing tools became stuck. They were backed off, leaving an additional 81-ft fish in the hole. The top of the fish was now at 4,935 ft, so the well would be sidetracked.

A kick-off plug was set and dressed to 4,660 ft. To resume drilling, the operator first tried to dry the hole by circulating from the casing shoe. The hole would not dry, so the 9/s-in. casing would need a retest. A packer was set at 4,125 ft. The casing would not test, and a leak in the casing was isolated above 135 ft.

A 60-arm casing caliper log and a magnetic thickness log were run but gave no clear indication of the problem. The operator decided to back off the 9/s-in. casing. While the casing slips were pulled, the casing parted in a pin at 121 ft. Two additional joints were backed off at 206 ft. Five joints of 9-/8-in. casing were then run and tested.

Examination of the failed pin from 121 ft indicated a fatigue crack had propagated in the connection prior to failure. The fatigue crack extended completely through the pipe wall below the root of the last engaged pin thread and approximately one-third around the circumference. Ductile fracture surfaces were at both ends of the fatigue crack and were a few inches long. These surfaces fractured when 183,000 lb were applied with the casing spear in the attempt to pull the casing slips.

Following the extension of the fatigue crack by these ductile fractures, the pin thread could no longer sustain the axial load and jumped out of the coupling because of the lack of remaining thread engagement. In addition to the failed joint, two other pins had fatigue cracks at the last engaged thread. The failed joint and both joints with partial fatigue cracks involved field-end connections (Fig. 1).

The flat brittle fracture surface associated with fatigue could be seen around most of the circumference of the crack. On both ends of the crack, a 45 ductile shear fracture occurred from the tension applied during the spear attempt on the casing.

WEST TEXAS

A string of 8 5/8-in. casing was run and cemented to 5,025 ft in a well. The slips were set with 120,000 lb total tension. A temperature survey located the top of cement at 3,400 ft. The 8 5/8-in. casing was a mixed string with 2,660 ft of 32 lb/ft, K-55 STC on bottom and 2,365 ft of 24 lb/ft, J-55 STC on top.

After the casing was tested and drilled out and new formation drilled, the kelly had signs of being polished from contact during rotation. A wear ring was installed in the wellhead to protect the slip area. The bit run was completed, and the bit was pulled with a total of 170 hr and 4,775 ft drilled. A new bit was run and a ledge was found at 262 ft. An impression block indicated the casing was parted.

The blowout preventer (BOP) was nippled down, and the 8/8-in. landing joint was speared. Six joints of casing came out of the hole. The sixth joint had a collar on bottom, meaning a mill-end connection had failed.

An overshot was run, and the pin was fished at 262 ft. This trip yielded four joints of casing that were free, leaving a new top of fish at 434 ft. A collar came out on the bottom of the fourth joint (tenth joint from surface) with a partial pin broken off in it looking down. The overshot was run again and engaged the casing. Tension was applied, and the casing was found in good condition. One additional joint of casing was backed off to leave a box looking up. Eleven new joints of S'/8-in., 24-lb/ft J-55 STC casing were run and screwed into the fish. The casing was set in 85,000-lb tension with a new set of slips.

The casing retrieved from the hole at 434 ft had a fatigue failure at the last engaged thread. Following this failure, drillstring/casing interaction caused rotation of a portion of the loose casing between the surface and the fatigue failure point. This rotation caused the connection at 262 ft to back out, as evidenced by those threads being in good condition and not deformed by jumpout or thread shear failure mechanisms.

The connection at 262 ft had simply unscrewed from the right-hand rotation of the casing below that point. Magnetic inspection of the joints pulled from the well revealed a second joint (from 393 ft, one joint above the fatigue failure) also had a fatigue crack. That crack was 4 in. around the circumference and, like the failure, was on the mill-end of the connection.

The 8 5/8-in. landing joint had severe wear from interaction with the drillstring. A hole was worn through the casing, and the slips were worn. Visible wear was noted on the kelly within 24 hr of drilling out the 8 5/8-in. shoe. The kelly was bent and contributed to excessive drillstring/casing interaction, as indicated by the amount of wear in the wellhead area.

FATIGUE AND STRESS

The following overview of fatigue and stresses in a connection help explain why the 8-rd thread is susceptible to fatigue failure.

The fatigue endurance of a material is frequently depicted in terms of number of stress cycles to failure versus fluctuating stress amplitude. Fatigue endurance is actually more complex, and other factors play an important role in determining fatigue endurance. One important factor is mean stress, the constant stress level about which dynamic stress fluctuations occur.

Thread geometry, makeup position, and connection ovality are key variables in establishing mean stress levels of a casing connection. Superimposed on these makeup stresses are the axial tension and bending loads the string experiences when it is run into the well. The makeup stresses combined with the overall structural loads determine the mean stress levels in the casing and the connections about which dynamic fluctuations occur during drilling.

Fig. 2 shows the effect of increasing mean stress on the fatigue endurance of materials characteristic of tubular steels. As the mean stress (a,) increases, fatigue endurance, depicted in terms of cyclic stress amplitude (a.), and cycles to failure (N) decrease. Mean stress in a casing connection is determined by connection design (thread geometry), dimensional tolerances (for both pipe and threads), makeup condition, and position in the string.

One of the most definitive industry works on the mechanical behavior of the 8-rd connection was performed by Asbill, Pattillo, and Rogers in the early 1980s.6-8 Their work combined laboratory testing of 8-rd connections with detailed finite-element analysis. Fig. 3 shows a plot of a finite-element grid used in their analyses. These types of studies have led to significant insight into the behavior of threaded connections and identified means of improving their reliability.

The study indicated threads transfer tension loads between the pin and box (coupling) in a nonuniform manner. The last engaged thread on the pin is subjected to higher thread loads than any of the adjacent threads because it is the first thread to transfer load into the coupling (Fig. 4). This condition creates a detrimental stress environment for fatigue endurance.

In addition to the nonuniform transfer of load in the 8-rd threadform, there is a stress concentration at the root of the thread. Fig. 5 illustrates the contact pressure generated on the 8-rd threadform during various load conditions. These results predict a significant variation in thread contact pressure along the thread flank with high contact pressures near the root and crest of the thread flank where contact with the mating threadform begins and ends, respectively. The high thread-flank pressure near the root contributes to the stress concentration and ultimately to the poor fatigue resistance of the 8-rd connection.

This overview of connection load and stress behavior does not attempt to quantify the fatigue endurance of American Petroleum Institute (API) 8-rd connections. It does highlight, however, the general mechanics of the connection under load and illustrates why the last engaged thread of a constant pitch connection and the thread root of the API 8-rd connection are susceptible to fatigue failure.

DRILLING CONSIDERATIONS

Because of the inability to quantify mann, of the variables that influence casing fatigue incidents, developing design algorithms with current technology is not feasible. However, in considering a typical casing design process in which the casing is selected on the basis of maximum burst, collapse, and tension loads, several points are worth highlighting.

In addition to the standard design checks, it is critical to perform a buckling analysis of the string and to include the necessary landing measures in the design process. For drilling casings, this analysis is addressed by examining loads during deeper drilling which typically involve the following:

• Heavier internal mud weight because of drilling into higher-pressured formations

• Thermal increases in the well, the mud, and the uncemented casing interval

• Degradation of annular fluid above the top of cement (the annulus becomes exposed to pore pressure as solids fall out of the mud and the effective density of the annular fluid decreases).

In normal operations, these three mechanisms drive the tendency for intermediate drilling casing to buckle. They can be easily analyzed by either a robust casing design system or a direct hand calculation. 9-10 In any case, it is imperative that a landing analysis be performed and buckling eliminated.

When slip-type hangers are used, both cement volume and landing tension can be used to eliminate buckling. If mandrel or landing ring suspension systems are used where no additional tension can be applied to the casing, cement must be sufficient so that buckling of the casing during deeper drilling is prevented. Internal pressure can be held on a casing string after it is landed while the cement sets to mitigate later buckling.

In addition to eliminating buckling, axial tension in a casing string increases lateral stiffness with regard to bending from drillstring impact. (This phenomenon is analogous to the tightening of a guitar string.) This tension also increases the mean stress level of the joints in the string. It is not currently known whether fatigue considerations would favor greatly adding tension to the string to maximize lateral stiffness (and hence minimize dynamic stress fluctuations) or limiting tension to that necessary to avoid buckling (and hence allow higher dynamic stresses about a lower mean stress). The current recommendation is to use a landing tension based on the elimination of buckling plus a reasonable safety margin.

At the cssing design stage, the landing analysis is an approximation based on planned parameters. This analysis is important for it includes the effect of landing tension on the structural integrity of the string.

Following running and cementing of the casing in a hole section, the landing calculations should be revised to account for variations between planned and actual conditions. These variations could include changes to the casing setting depth, cement volumes, top of cement (with regard to lost circulation, excessive washouts, etc.), temperature profiles, mud weight, and pore pressures. Where necessary, conservative assumptions should be made to ensure that landing tension is sufficient. The refined calculation should be clearly communicated to the field personnel to ensure that landing procedures are appropriate and properly executed.

BUTTRESS CONNECTION

Buttress thread casing (BTC) connections have higher tension capacities than 8-rd (STC and LTC) and are less susceptible to jumpout failure because of their improved threadform for sustaining axial loads. What is less defined, but intuitive from the geometry and implied from field data, is that the BTC thread has superior fatigue characteristics compared to 8-rd. This conclusion is proposed because ARCO has experienced no fatigue failures with BTC connections, and to the authors' knowledge, other operators have not experienced these failures with BTC connections.

Despite these advantages, 8-rd connections are prevalent in many operating regions, primarily because of lower cost. The estimated cost differential between 8-rd and BTC varies depending on the source. Some casing design systems assume a premium for BTC of about 7-11% greater than STC and 67% greater than LTC.

Recent work by ARCO, however, has found that BTC connections are available for premiums as low as 1-2% greater than STC or LTC prices. This low cost premium for BTC and its structural advantages provide strong incentive for increased BTC usage relative to that for 8-rd.

CEMENTING

"Tack cementing," or cementing a minimal length of the casing near the shoe, is a common practice and is influenced by a number of factors. These factors include the desire to minimize cement costs, ability to cut and retrieve casing on exploration wells, and the need to sustain annular communication with the open hole for cuttings injection. Although each reason is valid, it must also be realized that leaving long sections of casing uncemented creates an opportunity for the drillstring/casing interaction to result in fatigue failure of the casing.

The cement obviously does not affect the magnitude of contact nor impact from the drillstring into the casing. However, the cement sheath prevents motion of the casing and absorbs or transfers those loads to other casing members or.the formation, thereby protecting the casing from dynamic stresses. These casing integrity factors should be considered in planning cementing programs.

DIMENSIONS

The dimensions of 8-rd connections are obviously a critical determinant of the stress levels in the connection. Some poorly machined 8-rd connections have developed cracks in the thread root near the last engaged thread.

Fig. 6 shows the basic geometry of the 8-rd connection. The cause for these cracks is believed to be a premature transition from the connection taper over the L2 dimension to the 12 runout. A premature transition to the runout angle results in excessive thread interference in a made-up connection when coupling threads mate with incomplete thread roots on the pin.

API is currently implementing procedures to improve dimensional specification and gauging of the 8-rd product so that these errors can be corrected.

The stress levels generated in such poorly machined connections were reportedly high enough to have caused microcracks in the thread roots immediately upon makeup. Obviously, the presence of such cracks at makeup is very detrimental to fatigue life in service.

Although these dimensional problems are not believed to have been an issue with the failures discussed in this article, readers should be aware of the potential for problems in the machining of 8-rd threads and be prepared to investigate them should a failure occur.

OVALITY

There are currently no direct industry limits on the degree of ovality (out of roundness) which can be present in casing or the connection. Indirectly, casing ovality is controlled by tolerances on outer diameter (OD) and drift. Casing OD tolerances are currently +1% and - 0.5 %, providing for a theoretical allowable ovality of 1.5%. This value is a severe degree of ovality and beyond that normally observed in practice.

Casing ovality aggravates stresses in the connection region because some of the casing ovality will inevitably appear in the connection itself. Connection ovality is a complex issue involving gauging methods, procedures, and other factors. To resolve ovality measurements, options such as API ring/plug standoff gauges versus proprietary pitch diameter gauges and gauging in the threading machine versus on the rack must be addressed.

Ovality in the connection (both pin and box) must be alleviated through deformation during makeup. This deformation generates additional stresses which are superimposed on those stresses generated during the makeup of an axisymmetric connection.

A number of companies have adopted certain allowances on ovality as part of their inspection criteria for tubular goods. Table 2 shows example allowances, which are considered liberal.

Obviously, tighter tolerances would improve the structural integrity of the string by decreasing stress levels in the connection and increasing pipe body strength, particularly for collapse.

Users should be familiar with ovality issues and implement appropriate inspection practices to minimize the ovality present in drilling casings.

MAKEUP CONTROL

Companies should adhere to API guidelines for all mill-end makeups. API guidelines specify the dimensional requirements for proper power-tight makeup.

Makeup torque is a secondary factor and should not be used as the primary control on makeup. With access to the inside of the coupling during mill-end makeup, direct measurements can be made on the penetration of the mill-end pin into the coupling. Because this measurement can be performed quickly, these measurements should be made on 100% of the connections, and the makeup torque to achieve this position should be recorded for each joint. Following the coupling installation for a lot of pipe, histograms of mill-end makeup torque should be prepared.

These histograms are important:

• They provide a consistency check on the mill-end makeups and allow any joints which required excessively high or low makeup torque to be identified and removed.

• they provide a characteristic normal distribution curve for makeup torques which can be used as a guideline in the field.

These guidelines are valuable since control of makeup in the field can be more difficult than in the mill because the position of the pin is less visible. In the field, this can result in a tendency to run casing on the basis of torque and not position, which can introduce significant errors if done improperly.

The mill-end statistics provide a means of running casing in the field on the basis of torque, but position still needs to be verified because of potential variations in required torque from horizontal mill-end makeups to vertical field makeups. In addition, other factors including temperature, makeup speed, and thread compound can vary. Thus, position-based makeup control is recommended until enough of the string has been run to verify the mill-end torque statistics in the field.

DRILLSTRING/CASING INTERACTION

Measures should be taken to minimize drillstring and casing interaction. The derrick, rotary, and wellhead should all be in alignment and level.

In addition, personnel should be alert to signs of excessive drillstring and casing interaction. A ditch magnet can be used to detect casing wear and unusual increases in metal returns during drilling. Wear of the wellhead bushing can also help identify alignment problems. Bouncing of the kelly, excessive variations in rotary torque or speed, and any oscillatory or harmonic motion of the kelly, drill pipe, wellhead, or BOPs are all signs that drillstring dynamics are active.

Remedial measures can include changes to fundamental operating parameters of weight on bit and rotary speed, use of shock subs, rotary feedback systems, drilling mechanics measurement-while-drilling tools with accelerometers, and sophisticated surface-based dynamic monitoring systems.

Because of the high vibration damping provided by the drilling fluid, these drillstring dynamics issues become more critical for air, mist, or gas drilling. In these cases, drillstring dynamic phenomena will have a greater magnitude than if drilling fluid were in the hole dissipating the vibration energy.

RESULTS

• Fatigue failure of 8-rd casing can occur in wells where uncemented casing in a vertical orientation is subjected to drillstring/casing interaction.

• Fatigue failures are more likely in lighter weight casings such as 95/s in. 36 lb/ft (D/t ratio of 27.3) or 85/s in. 24 lb/ft (D/t ratio of 32.7) but are also possible in more common weights such as 95/8 in. 47 lb/ft (D/t ratio of 20.4).

• Fatigue failures can occur in either mill-end or field-end connections, usually in the connection with the highest makeup torque. In some cases, both mill-end and field-end connections can be fatigued.

• Fatigue failures can occur in any grade of pipe.

• Fatigue failures occur in both the STC or LTC versions of the 8-rd threadform.

• A number of factors increase drillstring/casing interaction and can lead to fatigue failures in combination with other contributing causes.

  These factors include drilling inside buckled casing, drilling with air or gas without having the beneficial vibration damping from a drilling mud, and drilling with misaligned or bent rotary equipment.

• The drillstring/casing interactions which can lead to fatigue failures are not necessarily excessive, beyond those in normal operations, nor detectable without special monitoring systems. Thus, many operations are at risk, given the correct combination of variables.

RECOMMENDATIONS

• Because both STC and LTC threads are subject to these fatigue failures, consideration should be given to the use of BTC threads in wells where the contributory factors to these failures are present. The extra cost of BTC over STC or LTC is small.

• If 8-rd threads are used in applications where fatigue is a possibility, the dimensional characteristics of the casing should be verified. The standard inspection should include measuring the casing and connection ovalities and checking the L2 dimensional transition into the 12 runout.

• Landing procedures should be included as part of the casing design process and refined prior to landing based on actual well conditions. Buckling must be eliminated in all drilling casings.

• In addition to proper landing tension, the uncemented interval should be minimized to increase the structural support provided to the casing by the cement sheath.

• Adequate quality assurance should be provided for both mill-end and field-end makeups. The use of statistical histograms of mill-end torque to power-tight position as a guideline for field makeup is recommended, but this technique should also be validated with observed field torques to achieve proper position.

• Drillstring/casing interaction should be minimized by leveling and aligning rotary equipment, avoiding the use of bent equipment, and using drill pipe rubbers if necessary.

ACKNOWLEDGMENT

The authors would like to thank ARCO British Ltd., ARCO Oil & Gas Co., and ARCO Exploration & Production Technology for permission to publish this article.

REFERENCES

1. Plaquin, M., Bourdon, F., Mandry, P., Puisais, X., and Kaluszynski, M., "Fast and Safe Method for Small Diameter Pipe Laying in the Deep Sea," presented at the Deep Offshore Technology Conference, Malta, October 1983.

2. Brown, D., and Bartle, M., "The Cause and Cure of Vibration-Induced Failure of Drill Casing in High Tidal Currents," paper No. 338, presented at the European Petroleum Conference, London, Oct. 25-28, 1982.

3. Bruno, T.Y., "How to Prevent Transit Fatigue to Tubular Goods," Pipe Line Industry, July 1988, pp. 31-34.

4. "Fatigue Test Results of 7-in. 32# Premium Connection," Sumitomo Metal Industries, technical paper G485, March 1992.

5. Texter, H.G., "Oil-Well Casing and Tubing Troubles," presented at the Spring Meeting of the Southwestern District of the American Petroleum Institute, Division of Production, New Orleans, March 1955.

6. Asbill, W.T., Pattillo, P.D., and Rogers, W.M., "Investigation of API 8 Round Casing Connection Performance-Part 1: Introduction and Method of Analysis," Journal of Energy Resources Technology, Vol. 106, March 1984, pp. 130-36,

7. Asbill, W.T., Pattillo, P.D., and Rogers, W.M., "Investigation of API 8 Round Casing Connection Performance-Part II: Stresses and Criteria," journal of Energy Resources Technology, Vol. 106, March 1984, pp. 137-43.

8. Asbill, W.T., Pattillo, P.D., and Rogers, W.M., "Investigation of API 8 Round Casing Connection Performance - Part III: Sealability and Torque," Journal of Energy Resources Technology, Vol. 106, March 1984, pp. 144-;4.

9. Kluementich, E.F., and Jellison, M.J., "A Service Life Model for Casing Strings," SPE Drilling Engineering, April 1986, pp. 141-52.

10. Craft, B.C., Holden, W.R., and Graves, E.D. Jr., Well Design - Drilling and Production, Prentice-Hall Inc., Englewood Cliffs, N.J., 1962.

11. American Petroleum Institute Specification Standard 5B, "Specification for Threading, Gaging, and Thread Inspection of Casing, Tubing, and Line Pipe Threads," Thirteenth Edition, May 31, 1988.

Copyright 1994 Oil & Gas Journal. All Rights Reserved.