Field Correlation Between Internal Taper Length and Tube Failures in 4.5-in., 16.60E, IEU Drillpipe

Abstract

Summary Noble Drilling Corp. has become increasingly concerned with premature failures in 4.5-in. [11.4-cm], 16–60 Grade E, IEU drillpipe tubes that meet industry requirements for metal structure, hardness, strength, and chemical composition and that show no signs of physical abuse. Our study of the pipe structure showed that a high percentage of tube failures were in tubes with short internal taper lengths. We examined a string of drillpipe with 18 tube cracks in the first 2,800 rotating hours in an area where 13,000 to 18,000 hours is a common life expectancy and studied drillstrings produced by three other manufacturers. We found that 80% of our failures occur within 10 to 12 in. [25 to 30 cm] of the internal taper, that the taper lengths are shorter among the failed joints than among a general population of drillpipe, and that these short taper lengths are not common to just one manufacturer or nation. We did not prove a causal relationship, but we did establish a correlation strong enough to warrant further investigation of the phenomenon and the establishment of a method that uses internal taper length as a criterion to inspect and purchase drillpipe. Introduction During our continuous study of drillstring failures, we have downgraded pipe because of necking and slip scarring, as well as wear and corrosive agent attack. We also have downgraded a considerable amount of 4.5-in. [11.4-cm], 16.60E, IEU drillpipe because of premature failures i.e., drillstring wall fractures that occur before half the normal pipe life expectancy has been achieved. In some operating areas we can estimate the average pipe life expectancy and cost of running the drillpipe. But even if the life expectancy goal is reached, the drillstring is the most expensive item on a drilling rig. When the drillpipe lasts only one-sixth of the normal operating life, there is indeed cause for alarm. The situation becomes even more critical because oilfield tubular inventories have been reduced by as much as 80% since 1981, and prices have begun to rise. This is not an isolated problem. Our personal communications concerning 4.5-in. [11.4-cm], 16.60E drillpipe include reports from one user with 24 failures in the first 500 rotating hours on a drillstring and from another user that averaged two failures per day out of a dozen strings in use. The Permian Basin chapter of the IADC has expressed concern and has formed a committee to investigate the problem. Also, the IADC and API established a joint data base to gather hard data to define the problem more clearly. We sent samples of our premature failures to independent laboratories for analysis. The results were almost always the same - the metal met all industry standards for structure, chemical composition, hardness, and strength, and the drillpipe showed no sign of physical abuse. The residual tests were not completely conclusive, but no signs of corrosive chemicals were found. The final conclusion was simply fatigue failure with a single crack penetrating the pipe wall. Detailed, professional inspections of the remaining drillpipe showed no sign of any other cracks or defects, but wall fractures continued to appear on each well, which necessitated that entire strings be laid down. Drillpipe has a certain resistance to stress and fails when the stress being applied is greater than design limits or when the stress resistance is lowered by stress risers (scars, corrosion pits, and abrasion) or by chemical embrittlement. On the basis of our analyses, we assumed that the drillpipe was designed and produced to industry standards and ruled out many obvious stress concentration factors in most of our premature failures. We looked at the structure of the tube for a stress riser, particularly at the upset internal taper, where the wall thickness changes most abruptly. We limited our study to 4.5-in. [11.4-cm], 16.60E, IEU drillpipe because it was our most commonly used pipe and the source of by far the largest number of failures. Definitions and Standards We standardized our measurements to the tube failures taken from the base of the tool-joint taper (Point A, Fig. 1). This point was selected to eliminate the difference in length between the box and pin to negate the effect of any tool-joint recutting. The weld is located 1 1/2 in. [3.8 cm] from this point; the external upset taper begins at a minimum of 3 in. [7.6 cm], and the internal taper, which has been measured at as much as 9 in. [22.9 cm] from Point A, starts at a minimum of 4 in. [10.2 cm]. These minimums are set by the API. Fig. 2 is a close-up of the internal taper. (The taper length is labeled L.) The external taper has a minimum length of 1 in. [2.5 cm], over which the wall thickness drops a maximum of 11 to 12 %. The internal taper, however, is the site of an almost 50% drop in wall thickness, with no minimum length set. Table 1 shows that 4.5-in. [11.4-cm], 16.60E, IEU drillpipe is the only type of Grade E drillpipe that does not have a minimum standard. If this wall-thickness drop were uniform over a 2-in. [5.1-cm] length, the standard minimum for all other types of pipe, the taper angle would be about 9.5 degrees. Unfortunately, the drop is not uniform because about 80 % of the wall thickness changes in about 20 % of the taper length. We have actually measured a taper angle greater than 60 degrees. Finite-element analysis models have been used to calculate the cumulative fatigue effects of hoop, radial, tensile, and torsional stresses. This type of mathematical treatment depends on the use of many beam elements and variables. Fortunately, some drillstring forces and motions can be quantified by actual downhole measurements, which reduce the uncertainties and number of assumptions that must be made. The stress concentration factors are also substantially affected by the radius of curvature in the internal taper. While rigorous mathematical treatment of forces and angles is an integral part of a total investigation, the purpose of this study was to evaluate the field evidence to understand a large and possibly growing problem. Test Procedure Pipe Sampling. We limited our tubing-failure samples to 4.5-in. [11.4-cm], 16.60E, IEU drillpipes that were either being used or had been used since 1979 in the Williston basin of North Dakota and Montana. All tubing failures had occurred downhole on wells averaging 10,300 ft [3140 m] in depth with maximum vertical deviations of 2.5 degrees. The mud systems were normally saturated salt with pH ranges of 6.5 to 7.0. Our first goal was to determine whether there was at least a tube-location relationship between the tapers and the failures. The first sample group (taken from our records) consisted of 247 tubing failures from 17 drillstrings. The measurements from the tapers to the holes or cracks were taken and recorded at the time of the failures. The next goal was to establish whether there was a relationship between the taper length and failure. SPEDE P. 58^