Dynamic Loading of Drillpipe During Tripping

Abstract

Summary. Dynamic phenomena when a stand of drillpipe is run into the hole may result in pipe failure, either through exceeding the yield or through kicking the pipe off the slips. Two computer programs based on longitudinal wave propagation have been developed. Results are presented and recommendations are made. Introduction Sec. 12 of the API RP 7G pertains to the dynamic loading of drill-pipe during tripping. The old text. based on work by Vreeland, was considered to be not sufficiently specific. A questionnaire, pertaining to drillpipe failures during tripping, was distributed at the March 1984 IADC Drilling Technology Conference. As a result, the API Committee on Standardization of Drilling and Servicing Equipment expressed a wish for a more specific text. A new text of Sec. 12, resulting from studies by an API work group and a research project with Amoco Production Co.. has been included in the 1987 edition of the RP 7G. This paper deals with research conducted up to June 1986 to prepare the new API text and after June 1986. Simplified System For explanatory purposes, consider an oversimplified system of Fig. la, consisting of drillpipe only with no drill collars, tool joints, or damping. The unit of time is taken as a one-way propagation (sound) travel in the pipe. Fig. 1b is a graph of imparted velocity at the surface. At Time 0, the velocity is increased suddenly from zero to a finite downward value, vs, and thereafter kept constant. The imparted velocity of Fig. 1b results in a dynamic force at the surface shown in Fig. 1c. A dynamic compression, d, appears at the surface. This compression remains constant until Time 2 - i.e., until the reflection from the bottom arrives at the surface. Then suddenly the compression d becomes a tension d until Time 4. Thereafter, the phenomenon repeats itself periodically, as shown in Fig. 1c. The velocity of sound in steel is 16,893 ft/sec [5149 m/s]. Therefore, if the pipe length were 16,893 ft [5149 m], then the numerals in Fig. 1c would be seconds. The period of oscillation would be 4 seconds. If the pipe length were 8,447 ft (16,893/2) [2574 m], then the period would be 2 seconds, i.e., the frequency of oscillations would double. Fig. 1c pertains to the dynamic force at the surface. Superimposing the static weight, W, on it yields Fig. 2. The tension varies between "b" and "c." From now on in this paper, the sum of the dynamic force and the static load will be called dynamic load. At Time 0, when the velocity vs is imparted at the surface, SF in Fig. la is infinitely small. In Fig. 1b, the velocity increases instantaneously from zero to a finite value-i.e.. with an apparent infinite acceleration. This is not so because the accelerated mass SF in Fig. 1a is infinitely small. Above F. the pipe is moving, and below F, it is still stationary. The front F is moving, downward with the velocity of sound. One might say that the Point S at the surface does not even feel how long the pipe is. The first information about its length is at Time, 2, when the front, F, after having reached the bottom. B, is reflected and returns to S. Suppose that in Fig. 1a the pipe is held in the slips that are released at Time 0. The plot of velocity at the surface (Point S) vs. time is shown in Fig. 3. At Time 0 the velocity at the surface instantaneously increases from 0 to 16 ft/sec [0 to 4.9 m/s]. This value is independent of the size and length of the pipe and of the presence or absence of drill collars and/or toot joints in the string. Suppose, however, that the length is 8,447 ft [2574 m] and that there are no drill collars or tool joints. Then, as shown in Fig. 3, the velocity of 16 ft/sec [4.9 m/s] remains constant until 1 second after the slips are released, at which time it increases by 32 ft/sec [9.8 m/s] to 16 + 32 = 48 ft/sec [ 14.7 m/s]. At each additional second thereafter, the velocity increases by 32 ft/sec [9.8 m/s]. The average acceleration (the slope of the dashed line in Fig. 3) is 32 ft/sec [9.8 m/s], the acceleration of gravity. Drillpipes With Collars and Damping But Without Tool Joints Two computer programs have been developed to investigate dynamic phenomena during tripping. The results presented here are for 41/2-in. [11.4-cm], 16.6-lb/ft drillpipe arid 61/2-in, [16.5-cm] -OD, 2 13/16-in. [7.1 -cm] -ID drill collars. The programs were used to investigate dynamic phenomena when going into rather than pulling out of the hole because the questionnaires filled out by drilling contractors indicated that failures were more prevalent when going into the hole. Fig. 4 pertains to adding a stand of drillpipe when 108 stands of pipe and 24 collars are already in the hole. The upper graph is the assumed idealized profile of velocity vs. time. During the first 2 seconds, the pipe is pulled up from the slips-i.e., moving upward. Then the pipe is dropped to its maximum velocity, v, which in Fig. 4 is about 8 ft/sec [2.4 m/s]. (The velocity at 1 second from the start is assumed to be -vmax/6.) The velocity of about 8 ft/sec [2.4 m/s] is maintained until about 12.4 seconds from the start. Then the pipe is decelerated in 1 second, at which time the velocity becomes vmax/3, and the slips are set. It has already been explained that imparting a finite velocity instantaneously results in a finite force. Similarly, setting the slips while the pipe is still moving results in a finite and generally acceptable force. However, such a procedure, although frequently used, is not recommended because it could result in slip damage to the pipe. In Fig. 4, the time of lowering the stand into the hole (running time) is about 13.5 seconds. The area under the curve (negative for the first 2 seconds) is 90 ft [27.4 m], which is the length of the stand. The lower graph in Fig. 4 is the axial force response vs. time, which is somewhat similar to Fig. 1c. In this figure, however, neither the period nor the amplitude of the oscillations is constant. Moreover, the shape of the curve changes from period to period. The amplitude of oscillations depends on damping. (Damping will be treated later.) For now, let us state that the system is slightly damped. In Fig. 4 the largest dynamic tensions, both before (Point A) and after (Point B) the slips are set are above the yield and the pipe could fail. Fig. 4 is for a running time of 13.468 seconds. For running times between 11 and 21 seconds, the computer analyzed 173 cases similar to that in Fig. 4. For some of them, the largest dynamic loads are greater than that in Fig. 4. This figure is presented here because the dynamic load after the slips are set is the smallest, namely about 25,000 lbf [111 × 10 N], which is still positive. JPT P. 975^