You must log in to edit PetroWiki. Help with editing

Content of PetroWiki is intended for personal use only and to supplement, not replace, engineering judgment. SPE disclaims any and all liability for your use of such content. More information

Coiled tubing fatigue

Jump to navigation Jump to search

Coiled tubing (CT) well intervention and drilling operations require that the continuous-length tube be subjected to repeated deployment and retrieval cycles during its working life. The tubing stored on a service reel is deployed into the wellbore to the designated depth and then retrieved back onto the service reel. The working life of the CT may be defined as the duration of service for the continuous-tubing string when subjected to the following conditions:

  • Bend-cycle fatigue
  • Internal pressure loading
  • Applied axial loading
  • Corrosion
  • Mechanical damage

All of the aforementioned items act on the tube body to some degree during any CT service and contribute to the eventual mechanical failure of the tubing. To ensure safe and reliable well intervention and drilling operations, the user must understand the unique behavior of CT to minimize the possibility of tubing failure. Numerous decisions must be made throughout the working life of a CT string to maximize the remaining life. From this approach, the decision to retire the tubing must be made on the basis of current tube conditions, service history, and the anticipated service loading.

Bend-cycle fatigue

Fatigue is generally considered to be the single major factor in determining the working life of CT. The deployment and retrieval of the continuous-length tubing string require that the tube be subjected to repeated bending and straightening events, commonly referred to as “bend-cycling.”

The amount of strain imposed upon the tube body during the bend-cycling process is considered to be enormous, in many cases on the order of 2 to 3%. When subjecting the CT to this type of fatigue cycling, the stress and/or strain fluctuations to failure may be estimated using conventional axial fatigue life prediction approaches.[1]

However, when the bend-cycling process is coupled with internal tube pressure loading, conventional multiaxial life prediction approaches cannot accurately predict CT behavior. Numerous tests performed have confirmed the fact that bend-cycling CT with internal pressure loading dramatically reduces the fatigue life of the tube when compared to the cycle life of unpressurized tubing. This happens despite the fact that the tangential (hoop) stresses imposed by typical “high” pressures in CT service are on the order of only 50 to 60% of nominal yield strength . However, when combined with fluctuating axial strain ranges on the order of 2 to 3%, significant cyclic ratcheting (ballooning) occurs, and fatigue damage development is accelerated. Outside of the CT industry, there are essentially no other applications involving steel alloys in which cyclic loading of this magnitude is intentionally imposed and expected to survive the prescribed service.[2]

Stress and strain in HSLA steel

To better understand the abuse imposed upon CT during normal service operations, a brief review of the relationship between stress and strain in high-strength low-alloy (HSLA) steel is given next. Fig. 1 is a typical stress-strain curve for HSLA steels. The axial stress, α, is plotted along the Y-axis, and the corresponding axial strain, ε, is plotted along the X-axis. As stress is applied to the steel material, a corresponding strain develops as defined by Hooke’s law that states that the amount of stress is equal to the amount of strain, multiplied by the modulus of elasticity of the material. This relationship is graphically represented as the line segment O-A in Fig. 1, where the modulus of elasticity defines the slope. The stress at Point A is referred to as the “proportionality limit,” also referred to as the “elastic limit.” As long as the stress levels within the steel are held below the elastic limit, the strains are also considered to be elastic, and no permanent deformation will occur.

However, as increased stress is applied to the steel, the elastic limit is exceeded and will reach Point B, which is termed the "yield point". The yield point defines the yield strength , Sy, or the stress that causes the initiation of plastic strain in the material. Once the yield point is reached, permanent deformation occurs, and plastic strain is developed as the material begins to elongate permanently.

To determine the yield strength of alloy steel materials in a consistent manner, a standard 0.2% offset plastic strain value was adopted to locate the yield point on the stress-strain curve. This is shown as the dotted line "AB-X.2%." Loading into the plastic region to Point P, followed by unloading to Point O, Line P-O is usually believed to also define the slope of the modulus of elasticity and intersect the X-axis to represent the amount of plastic strain resulting from the deformation. However, it has been shown that plastic deformation tends to reduce the modulus of the elastic unloading curves by as much as 15 to 20% for loading typically seen in CT well intervention and drilling services.

By applying additional stress, we will reach Point D, which is referred to as the ultimate tensile strength of the material. Once Point D is reached, the material will suffer a separation failure.

The amount of stress experienced by the HSLA material can be fully appreciated when considering the degree of bending the tube must undergo during conventional service activities. The radii of typical CT service reels and tubing guide arches are significantly less than the yield radius of curvature (RY) for any given size of CT listed. When the tubing is bent and wound onto the service reel or storage spool, the HSLA steel is yielded and becomes plastically deformed. This plastic deformation event is similar to the graphical illustration seen in Fig. 1 as Curve O-P. When the amount of stress applied to the tubing is relaxed, the strain remaining is permanent and has a value represented by the extension of Line P-O'.

The bend event sequence

The fatigue imparted to the CT material during normal service operations is the result of bending the continuous-length tubing beyond its elastic limit and forcing the material into plastic deformation. Fig. 2 illustrates the typical operating sequence whereby bend cycles are imposed on the CT during deployment and retrieval. For this illustration, the initial state of the tubing will be in the "as wrapped" condition on the service reel. The bend event sequence is described next.

  • The tubing is pulled off the service reel by the injector. The hydraulic motor on the reel provides resistance to the pull of the injector, placing the tubing in tension. The tension applied is typically limited to the amount needed to bend the tubing over the tubing guide arch and maintain control of the tube during deployment. Therefore, the tubing is straightened out somewhat, constituting the first bend event.
  • Once the CT reaches the tubing guide arch, the tubing is bent over the prescribed radius. This event constitutes the second bend event experienced during deployment.
  • As the tubing travels over the tubing guide arch, the tubing is returned to the straightened orientation before entering the gripper blocks in the injector (third bending event).

These three bending and straightening events are repeated in reverse as the tubing is extracted from the well, resulting in a total of six bending events commonly described as a “trip.”

In relative terms, the smaller the bending radius, the larger the value of bending strain induced into the tube segment. This is seen as the more severe bending strains imposed when the CT is cycled over the service reel, as compared to the less severe bending strains imposed when bending occurs over the tubing guide arch. Therefore, during a complete service trip, the tubing will experience two bending events inducing relatively higher bending strain and four bending events inducing relatively lower bending strain. Note that the pairing of two bending events with equivalent strain magnitudes is defined as a “bend cycle.” Therefore, within each CT service trip, there are two relatively low-strain bend cycles and one relatively high-strain bend cycle.

In general, for CT well-intervention and drilling operations, the number of trips assigned to the tubing is not the same over the entire length of the tubing string. For example, during routine service work, it is common to periodically stop deployment and reverse the motion of the CT string, retrieving a prescribed length of tubing back out of the wellbore to check weight and drag of the tubing. The intermediate segments of the tubing string, which were subjected to these “drag checks,” will have a greater number of trips allocated than that which is allocated to the remainder of the tubing string. In addition, where a prescribed service program requires repeated bend cycling over a specified segment of the tubing string, this segment will have an accumulation of fatigue damage that is significantly greater than the fatigue damage subjected to the remainder of the tubing-string length.

The discussion on CT bend cycling makes apparent that most of the material fatigue occurs at the reel and tubing guide arch, with very little, if any, within the wellbore. In CT services, the plastic deformation of the tube material that results in the cumulative damage is defined as “ultralow cycle fatigue.” The loading in ultralow cycle fatigue is plastic, and failure of the material generally occurs within 2,000 bend cycles.

The occurrence of microcracks in CT operations

A consequence of ultralow cycle fatigue in CT operations is the eventual formation of microcracks in the tube body. With continued bend cycling, the cracks propagate through the tubing wall until the crack establishes full penetration from one side of the wall to the other. Once the crack has fully translated through the tube body wall, pressure integrity is lost. Typically, the initial size of the crack is very small, making detection of this type of body damage very difficult. Furthermore, cracks tend to initiate on the inner wall of the tubing, despite the fact that the bending strain is greater on the outer surface. In conditions in which high internal pressure is present, the hoop stress may cause the crack to instantaneously propagate along the circumference of the tube. In this condition, a major transverse crack will occur and may result in the mechanical separation of the tubing. For this reason, crack initiation is typically the criterion used to define “failure” within a CT section.

Bend cycle test and data

Loss of pressure integrity in any condition makes the CT string unfit for service. At present, the fatigue condition of a segment of tubing with an unknown bending history cannot be accurately measured nondestructively. Because fatigue appears to be a statistical phenomenon, variations can be expected in observed fatigue life for any sample, regardless of comparable bending history records.

Various types of bend-cycle fatigue test machines (Fig. 3) have been developed in an attempt to simulate the wellsite bendcycling of CT over a constant radius with internal pressure loading. These bend-cycle fatigue fixtures generate a statistically significant quantity of fatigue data and provide a means to estimate bend-cycle fatigue over a wide range of test conditions.

Numerous analyses of the trends recorded from the ever-increasing volumes of CT fatigue tests suggest that bend-cycle events imposed onto a given tubing specimen with high internal pressure (causing hoop stresses on the order of 30 to 60% of the yield strength ) accumulate fatigue damage at a much greater rate than bendcycles imposed with low internal pressure loading.

In addition, the magnitude of fatigue damage realized from a given bend-cycle event cannot be applied to working-life predictions in a linear fashion. Evidence from the volume of testing suggests that a given bend-cycle load applied later in the tube working life may cause greater fatigue damage than the equivalent bend-cycle load applied earlier in the tube working life. The bend-cycle fatigue data obtained from the test fixtures has also provided insights into several other areas of CT service behavior. These items are discussed next.

“Cyclic softening” of CT material

One significant consequence of repeated bend cycling of CT material is a phenomenon referred to as “cyclic softening,” which results in a reduction in material yield strength as the CT performs its prescribed service.[3] The magnitude of the cyclic softening realized in CT materials can be directly related to the amount of material strain imposed during service.

Standard strain-controlled, low-cycle fatigue testing is commonly performed on axial-test coupons of a specified material to evaluate its behavior when subjected to fluctuating strain conditions. These tests attempt to simulate the magnitude and intensity of the strain fluctuations anticipated during the prescribed service and are used to help predict the fatigue strength of the material at the test conditions. For steel CT products, similar types of stress-cycle tests have been performed to provide insight into fatigue life and performance when subjected to multiaxial plastic deformation events.

The strain-controlled fatigue-testing process is illustrated in Fig. 4, in which the CT material is cut into the designated test coupon geometry and installed into a fatigue test machine that subjects the material specimen to strain-controlled, axial load cycles between a maximum and minimum strain, γmax and γmin . In these tests, cyclic softening was observed because the peak stresses 3, 5, and 7 continually diminish from the initial peak at Point 1.

The stress/strain diagram shown in Fig. 4 illustrates the hysteresis effect observed during the strain-controlled axial-stress load-cycle tests. If plotted in terms of stress vs. time as shown schematically in Fig. 5, the coupons exhibit a transient stress response, in which the recorded peak stress values diminish incrementally relative to the previous value. This cyclic softening corresponds to a reduction in the material yield strength . Eventually, stable hysteresis loops form, reaching a condition described as "cyclic stabilization" of the material. In the conventional low-cycle fatigue tests conducted on the material coupons, cyclic stabilization appears to occur when the accumulated fatigue damage is within the range of 20 to 40% of expended cycle life.

The material strain imposed during bend cycling of full-body CT specimens is considered to be at the upper limit of the low-cycle fatigue regime, especially because this bending can combine with pressure to cause lives on the order of fewer than 20 cycles. As previously discussed, the category that best describes the behavior of CT bend-cycling is that of “ultralow cycle fatigue.” At the strain levels realized during full-body CT bendcycling over typical bend radii, the bulk of the material softening can be seen within the first few load cycles, with continued material softening occurring over the remainder of the tube cycle life. For full-body CT bendcycling, truly stable stress responses may not be achieved, and cyclic stabilization may never be truly realized.

The importance of this discussion becomes apparent when considering that the strain range experienced in CT services varies significantly from point to point around the circumference of the tube body. At positions between the points of maximum stress (i.e., the top and bottom of the tube body) and the neutral axis, the intermediate strain ranges and transient material softening will most likely occur throughout the entire working life of the tube. As a result, material strength and performance properties can be expected to vary relative to the position around the circumference of the tube body. In some CT material samples tested, the reduction in yield strength was found to be as much as 10 to 20% of the parent material yield strength .

In actual service, the neutral bending axis on the CT when deployed into the wellbore will most likely be in a different orientation on the tube circumference from when it is retrieved from the wellbore. This change in neutral bending axis may be the result of relaxation of residual stresses in the tube when axial tensile forces are applied during deployment into the well or because of some other phenomena.

During the extraction process, the tubing is retrieved into the injector gripper blocks with the previous neutral axis rotated slightly off alignment. Once the tube is secured into its “new” orientation relative to the gripper blocks, a new neutral axis will be defined as the tubing is bent over the tubing guide arch. This repositioning of the neutral axis during well intervention and drilling operations is believed to distribute the high-strain bend cycles around the circumference, creating a more uniform accumulation of fatigue damage within the tube.

Diametral growth

When subjected to plastic deformation because of bendcycling with internal-pressure loading, the diameter of the CT tends to grow or “balloon.” Even when the internal pressure loading is well below the yielding stress of the material, the tube body is subjected to hoop and radial stresses that cause the material to grow macroscopically in diameter and to decrease in wall thickness (Fig. 6). In uncontrolled fatigue tests, diametral growth exceeding 30% has been observed. The primary factors influencing diametral growth are material properties, bend radius, internal-pressure loading, tube outside diameter (OD), and tube wall thickness.

Interaction with surface handling and pressure-control equipment

One major concern for diametral growth is the interaction with surface handling and pressure-control equipment. The injector gripper block loading on CT usually has an impact on the tube geometry, and the effect tends to vary according to variables such as:

  • The magnitude of gripper block normal force
  • Block geometry and wear
  • CT geometry, internal pressure, and material type

Most conventional counter-rotating chain injectors have gripper blocks that are machined to fit the OD of the specified size of tubing. When CT experiences diametral growth, the increase in tube size creates a nonsymmetrical loading condition, concentrating the normal force load at contact points on the edges of the gripper block. These focused stress concentrations induce additional damage into the tube body and result in added tube deformation.

Some injector manufacturers have made adjustments to the design of the gripper blocks in an attempt to minimize this additional damage to the tube. The segmented-type and “variable”-type gripper-block designs appear to better accommodate increases in tube body growth. However, as the diametral growth of the tube increases, the two-point contact loading of the gripper blocks will induce lines of concentrated stress on the tube body. With this focused stress concentration, the gripper blocks will impart greater damage onto the tube at the normal force loads typical for properly fitted gripper blocks.

Another concern for diametral growth relates to the interaction with the pressure-control equipment. The stripper assembly contains brass bushings that are used to prevent extrusion of the elastomeric elements. These bushings have an internal diameter that is slightly larger than the specified OD of the tubing. If the actual CT diameter on any axis reaches or exceeds the internal diameter of the brass bushings, the CT will bind up within the bushings, resulting in surface damage to the tubing. Once this condition is reached, the CT may no longer pass through the stripper. To prevent this situation, limitations should be placed on the maximum allowable CT diameter.

Absolute diameter

The recommended method of defining maximum allowable growth of CT is “absolute diameter,” at which the CT is retired from service when the diameter reaches a given value greater than its specified dimension. The limit typically used is 0.050 in. and can be considered valid for all CT sizes when the aforementioned type of stripper apparatus is used.

Observations of bend-cycle fatigue testing directly relating to diametral growth in CT are listed next.

  • The growth rate of the OD increases with increasing internal-pressure loading.
  • The diametral growth of larger-diameter CT as a percentage of specified diameter is greater than that of smaller-diameter CT.
  • CT specimens with higher material-yield-strength demonstrate less diametral growth than lower material-yield-strength specimens.
  • Mechanical limitations of surface handling and pressure-control equipment tolerances for allowable diametral growth restrict the effective working life of CT in high-pressure service to only a fraction of the projected fatigue life.

Differential wall thinning

As a consequence of diametral growth, CT experiences differential wall thinning. If we assume that the cross-sectional area of the tube body is constant, then as the diameter grows, the redistribution of material causes the wall of the tube to get thinner. The concept of differential thinning is also illustrated in Fig. 6.

As the tube is bend cycled about the neutral axis, the top and bottom of the tube are subjected to the highest stress concentrations and subsequently experience the greatest amount of thinning. Note that the wall thickness at the neutral axis (where no bending stress occurs) remains at the initial wall thickness. Therefore, the wall-thinning process is not uniform. Although the severe ballooning and localized wall-thickness reduction are typically observed on samples tested in the laboratory, such gross geometric changes are rarely seen in the field. This is because of factors such as tubing rotation (which causes the wall thinning to occur more uniformly) and the ID of the stripper brass bushings (which limit the use of tubing having excessive diametral growth).

Although extensive diametral growth and wall thinning are observed in the lab, empirical data obtained from extensive full-scale field cycle testing conducted by Walker et al.[4] has shown that for a 3% increase in tube body diameter, the wall thickness tends to decrease approximately 7.5%, resulting in a loss of burst and collapse pressure rating of as much as 10%.

The mechanics behind the phenomena of diametral growth and wall thinning is referred to as transverse cyclic ratcheting, whereby material deforms permanently with each cycle of loading in directions transverse to the major fluctuating load directions. Conventional incremental plasticity theory has been shown to overpredict cyclic ratcheting in the typical CT loading regime. Therefore, refined plasticity models have been developed and used successfully to predict the behavior of CT samples in the lab and the field.

Tubing length extension

As a result of the stress/strain response exhibited by CT material (as seen in Figs. 1 and 4), the tubing enters the well with a significant distribution of residual stress. The residual stresses vary from tension on one side to compression on the other, at magnitudes equal to the material’s cyclic yield strength. This residual stress profile has a first-order influence on the load-deflection behavior of the tubing and tends to induce secondary deformation mechanisms which contribute to permanent increases in tube length. This condition is commonly described as “elongation,” but should not be confused with the API definition of elongation, which represents the change in specimen gauge length once ultimate tensile strength is exceeded during a destructive tensile test. To avoid confusion, the term “extension” is used in this text to refer to the observed permanent increase in CT length resulting from realignment of residual stresses where axial force loading is applied.

As with diametral growth, extension occurs despite the fact that the applied axial stress (in terms of load over cross-sectional area) is maintained to a value substantially below the material yield stress. The investigation of extension was prompted by observations in coil-tubing drilling (CTD) operations where the bottomhole assembly (BHA) could not return to maximum deployed depth after the string was retrieved to surface and then redeployed to depth (referenced by painted points on the reel). The reported “stack-out” depth was several feet higher than the previously reached depth, in excess of calculated corrections for tube helix reorientation within the well. Extraction and subsequent redeployment of this CT string found the stack-out depth several feet higher than the previous “stack-out” depth, prompting concern for accurate depth control. Other onsite observations found that the amount of permanent tube length increase was greater for new tubing strings as compared to “seasoned” tubing strings.

Detailed studies have documented that extension occurs during high-pressure/constant-pressure bend cycling because of the constant axial stress caused by pressure loading on the end caps. When higher tensile loads are applied to the tube immediately after bend cycling has occurred, the section of the tube wall with residual tensile stresses equal to the yield strength cannot sustain any more axial stress. Therefore, the residual stresses within the tube body are redistributed to reach equilibrium, resulting in a permanent extension of tube length.

The intensity of the residual stress distribution in the tube body is directly related to the imposed strain, which suggests that the amount of extension for a given axial load increases as the bend radius is reduced. Therefore, the location of maximum tube-length extension within the string will typically be in the region where the smallest bend radii and highest axial tensile loads are imposed. This region of the tubing string is typically located near the core diameter of the service reel (reference end), with high axial loading resulting from the weight of the tubing suspended within the wellbore.

Multiaxial plasticity models have been developed to compute stresses and strains induced by the surface handling equipment (tubing guide arch and service reel), internal pressure loading, and axial loads experienced by the tube. These complex and sophisticated algorithms handle highly nonproportional loading and must accommodate changes within the tube cross section as well as axially along the entire length of the string. Therefore, the accuracy of the plasticity model predictions depends upon the proper mapping of imposed strains throughout the service life of the tubing string.

Mechanical damage

The bend-cycle fatigue life of a CT string is also sensitive to surface damage, such as:

  • Scars
  • Scratches
  • Gouges
  • Dents

These types of mechanical damage serve as localized stress-strain concentrations where repeated bend cycling can cause cracks to develop in the tube body.

Testing was performed by Quality Tubing to quantify the reduction in bend-cycle fatigue life for 1.750-in.-OD, 0.134-in.-wall-thickness tube samples, with and without surface damage present.[5] The 80-ksi-yield tubing samples were subjected to bend-cycle testing over an equivalent 72-in. bend radius with a 3,000-psig internal pressure. A baseline test was performed with an undamaged tube and survived approximately 725 bend cycles before loss of internal pressure (failure). Two samples were prepared with transverse notches cut to an approximate depth of 10% of wall thickness. One sample located the notch across the longitudinal weld seam, with the other sample locating the notch 180° from the longitudinal weld seam. Both of these samples survived approximately 120 bend cycles, yielding an 83% reduction in fatigue cycle life relative to baseline.

Two additional samples were prepared with the 10% wall thickness notches located at the same positions on each tube body, but these notches were ground out and polished smooth before being subjected to bend cycling. By removing the transverse notches, the tube samples survived approximately 590 bend cycles, yielding a 19% reduction in fatigue-cycle life relative to baseline. With the removal of the surface damage, the bend-cycle fatigue life increased from 120 to 590 cycles, a 390% improvement in survival life.

A recent study has noted that not only is the geometry of the defect important (length, width, depth and shape of the defect), but also the mechanism by which the defect was imposed into the tubing.[6] For instance, ball-nosed end mills were used to cut hemispherical defects into a set of CT samples. In another set of tubes, defects with identical geometries were imposed by pressing hard balls into the surface, rather than removing material by machining. When tested under identical constant amplitude bend cycling, the samples with the milled defects reduced the life of the samples by as much as 60% (relative to defect-free baseline samples), while the impressed defects caused virtually no life reduction.

Significant research is currently under way to quantify the influence of defects on CT life, as well as the effectiveness of repair strategies, such as simple removal by surface grinding. The research currently under way is addressing the influence of material grade and loading parameters such as:

  • Bending radii
  • Pressure
  • Tube geometry

An important component of the research is the incorporation of CT inspection technology. The most prevalent CT inspection technique is magnetic flux leakage. Although this technique has proven effective in detecting the presence of a flaw, existing technology provides no information about the severity of the flaw.

The research currently under way has made important steps[7] toward extracting quantitative information about the geometry of the defect, as well as its influence on fatigue. Over time, a continually growing database will facilitate the development of refined signal-processing algorithms that provide important feedback directly from magnetic flux leakage results.

Surface rippling

An additional phenomenon that occurs as a result of CT bend-cycle fatigue is described as “surface rippling.” It is common to find “ripples” developing on the top surface of the CT (relative to the neutral axis) when subjected to bend cycling in combination with high internal-pressure loading. On many samples examined, the period of the ripple formation has typically been observed to be approximately twice the tube diameter.

This condition has been observed to occur both in field service operations and with tube specimens tested in fatigue cycle fixtures. This consequence of bend cycling is attributed to the fact that CT does not yield continuously along its length as it is deformed. What occurs is more of a buckling phenomenon, where the outer fibers simply lengthen to a point where they buckle in compression when straightened and never quite return to a perfectly straight condition upon rewrapping. Rippling typically occurs relatively late in the fatigue life of the tubing. Therefore, if surface rippling is observed in the CT, it is recommended that the tube be immediately withdrawn from service.

Effects of bend cycle fatigue on welds

Welds of several types are fundamental to the manufacture of steel CT. Welds are a common concern because the bend-cycle fatigue life of certain welds can be significantly less than that of the parent tube material. As a result, the performance of welds is of critical importance to the working condition of the tubing string as a whole.

The longitudinal seam weld runs the entire length of the coiled tube and is created during the manufacturing process as the base skelp is formed into tubing in the mill. Problems with the longitudinal seam weld are usually detected at the mill during either the manufacturing process or the subsequent hydrostatic pressure testing procedure. However, failures along the seam have occurred during field services in high-pressure applications because of excessive triaxial stress loading.

During laboratory CT fatigue testing, it has become common practice to place the longitudinal weld seam along the curved bending mandrel or on the compressive side of the tube wall. In general, cracking occurs on the compressive side of the tubing at least as often as (if not more often than) it does on the tensile side of the tubing. This observation holds even for testing when the weld seam is placed along the neutral axis of the tubing.

In general, tests performed on the orientation of the longitudinal weld, with respect to the bending axis, have found no correlation of cycle fatigue life reduction because of longitudinal seam-weld damage. In the process of derating the CT because of bend cycling, the longitudinal weld is generally disregarded.

Skelp-end weld

The welding technique that joins base metal skelp is referred to as the skelp-end weld. The flat skelp is cut and welded at an approximate 45° angle, causing the weld to form a helical wrap around the tube girth when the strip is subsequently formed into a cylinder. The weld is typically made in ideal conditions (with full control over the geometry and weld penetration), resulting in a weld of high and consistent quality.

The skelp-end weld performs favorably in bend-cycle fatigue tests because of the:

  • Quality control
  • Treatment of the weld
  • Subsequent helical distribution of stress

Experimental results taken from bend-cycle fatigue tests comparing base tube samples to skelp-end welded samples found that the skelp-end weld survival can range from 80 to 90% of the base tube working life with relatively little variation from one weld to the next. This is well within the typical range of scatter exhibited by CT fatigue lives tested in the laboratory and the field.

Butt-weld/tube-to-tube weld

A butt-weld or tube-to-tube weld is a means for joining lengths of previously formed tubing. The two ends of the finished tube to be joined are cut square, slightly beveled, carefully aligned, and TIG- or MIG-welded (Tungsten inert gas or Metal inert gas) around the circumference. The resulting weld is oriented perpendicular to the CT axis, focusing the bending stress perpendicular to the tube axis (directly through the weldment). Because the weld is made from the tubing exterior, the quality of the weld penetration through the tubing wall is paramount. There are currently two types of tube-to-tube welds in use in CT services, which are discussed next.

Factory butt-weld

Before the development of the skelp-end weld manufacturing process, all CT strings were constructed by butt-welding several lengths of finished tubing together using an automated welder at the tubing mill. These welds were made in carefully controlled conditions, and the quality of these welds was generally very good. However, these types of welds remain significantly inferior to skelp-end welds, with bend-cycle fatigue life observed to be in the range of 40% of base pipe working life.

Field butt-weld

Once the CT is delivered to the field location, tubing sections which need to be joined are typically butt-welded by hand using a TIG or MIG technique. This “field repair” is frequently the only option available at the service vendor location. Manual butt-welds require a very high level of skill to have acceptable survival rates. As a result, field butt-welds are the most problematic of all CT welds. With the high likelihood of significant variations in quality of welds, the recommended practice based on experimental tests is to derate field butt-welds to approximately 20% of base tube working life.


  1. Tipton, S.M. and Newburn, D.A. 1996. Plasticity and Fatigue Damage Modeling of Severly Loaded Tubing. In Advances in Fatigue Lifetime Predictive Techniques
  2. Tipton, S.M. and Brown, P.A. 1994. Monitoring Coiled Tubing Fatigue Life. Presented at the International Conference on Coiled Tubing Technology, Houston, Texas, USA, 29-31 March.
  3. Tipton, S.M. 1997. Low-Cycle Fatigue Testing of Coiled Tubing Materials. Presented at the World Oil Coiled Tubing and Well Intervention Technology Conference, Houston, Texas, USA, 4-6 February.
  4. Walker, E.J. 1990. Collapse Tests Expand Coiled Tubing Uses. Oil Gas J. 88 (10): 56.
  5. Stanley, R.K. 1998. Results From NDE Inspections of Coiled Tubing. Presented at the SPE/ICoTA Coiled Tubing Roundtable, Houston, Texas, 15-16 March. SPE-46023-MS.
  6. Tipton, S.M., Moran, D.W., and Sorem, J.R. 2002. Quantifying the Influence of Surface Defects on Coiled Tubing Fatigue Resistance. Presented at the SPE/ICoTA Coiled Tubing Conference and Exhibition, Houston, Texas, 9–10 April. SPE-74827-MS.
  7. Moran, D.W., Chinsethagid, C., Sorem, J.R. et al. 2002. Challenges Facing the Development of Coiled Tubing Inspection Technology. Presented at the SPE/ICoTA Coiled Tubing Conference and Exhibition, Houston, Texas, 9–10 April. SPE-74835-MS.

Noteworthy papers in OnePetro

Sisak, W.J. and Crawford, D.J. 1994. The Fatigue Life of Coiled Tubing. Presented at the SPE/IADC Drilling Conference, Dallas, 15–18 February. SPE-27437-MS.

Tipton, S.M., Smalley, E., and VanArnam, D. 2012. Influence of a Straightener on Coiled Tubing Fatigue. Presented at the SPE/ICoTA Coiled Tubing & Well Intervention Conference and Exhibition, The Woodlands, Texas, USA, 27–28 March. SPE-154057-MS.

External links

See also

Predicting CT fatigue