AA ADE-11-NT TCE-31 Pre event Rot tary Shou uldered Connectio C on Failure es

Jim Douglas, Gag gemaker, LP

ht 2011, AADE Copyrigh This pap per was prepared for presentation p at the 201 11 AADE National Tec chnical Conference an nd Exhibition held at th he Hilton Houston Nor rth Hotel, Houston, Te exas, April 12-14, 2011 1. This conference wa as sponsor red by the American Association A of Drilling Engineers. E The inform mation presented in this paper does not refle ect any position, claim m or endorsement mad de or implied by the A American Association of o Drilling E Engineers, their officers or members. Ques stions concerning the content c of this paper should s be directed to th he individual(s) listed as author(s) of this wo ork.

Abst tract D Designers of th he original ro otary shoulder red connection n (RSC C) would never r have imagined d that their des signs would be e taken n to the extrem me depths and subjected to the t severity of f direct tional drilling as they are today. t Comp panies demand d super rior RSC perfo ormance as th hey run drilling tools longer r and w with higher for rces to maintai in schedules an nd reduce time e on lo ocation. Unfor rtunately, galle ed shoulders, swollen boxes s and f fatigue stress cracking c are co ostly connectio on failures that t are f fairly common n and an exp pected expens se by drilling g contra actors today. Th he root cause of most RSC C failures and damage is the e lack of thread int terference in the connectio on. The most t impor rtant thread el lement of an RSC R connectio on is the pitch h diame eter size, whi ich is not add dressed in the e current API I Speci ification 7-2 or r API RP7G. Consequently, C connectors are e still m manufactured using 92 year r old inspection technology y that p produces conne ections that do o not have full thread contact t or pro oper radial inte erference when n torqued toge ether. In order r for c connections to achieve the performance level required d today y, a new approach to gauging g and dimensio onal inspection n must be pursued an nd implemente ed for not only y new, but also o used connections. Th his paper ex xplains the im mportance of isolating the e conne ectors pitch diameters d from m the ring an nd plug gauge e stand doff and the oth her thread elem ments, making it a standalone e attribute measureme ent. By measu uring the pitch diameters, the e opera ators can directly control the t connector s size during g manu ufacturing as well w as ensure that t the thread d cones go into o interf ference properly when the e connections s are torqued d togeth her. Intro oduction R Rotary shoulder red connectors s are key elem ments of a drill l string g and are also among the mo ost frequent cau uses of failure. The most common n RSC failur re is fatigue which causes s shoul lder galling, box swelling and a stress crac cks; then these e problems lead to drill d string fai ilure and poss sible loss of a bottom m hole assemb bly. Drill strin ng and bottom hole assembly y (BHA A) component failures contin nue to afflict th he offshore oil l and g gas industry an nnually, involv ving direct and d consequential l costs reaching many y millions of dollars.

D Due to the incr rease in well d depths and the e complexity of o drilli ing operations, , the control o of drill string s serviceability is becom ming more a and more sig gnificant. B Because rotary y shoul ldered conne ections are o one of the most stressed comp ponents of th he drill string g and they are constantly y subje ected to both make-up torqu ue and to axia al and bending g loads s, their design n must be sy ystematically s scrutinized and inspe ected. Unfor rtunately, curr rent inspection n requirement ts utiliz zing nearly 10 00 year old ga auging equipm ment are not as a comp prehensive as t they must be. It t has been a sta andard practice to use ring a and plug gauges to siz ze both memb bers during th he design and manufacturing g produ uction process s. Not only are vital mea asurements no ot being g aptly scrutin nized when the e connection is s new, but used d pipe is never inspected to pinpo oint thread ele ement variation n whic ch can prevent t future failure es; used pipe i is only visually y inspe ected for stres ss cracks and other damage e. Therefore e, instea ad of a quick k fifty dollar c cleaning or ref facing expense e, opera ators face hund dreds or thous sands of dollars in outsourced mach hining and fr reight charges s. This reac ctive versus a proac ctive approach h must be chan nged and a ne ew approach to gaug ging and dimen nsional inspec ction for both new and used pipe must be pursued and implem mented. ed Drilling Ev vents Cost tly Unplanne R RSC failures an nd loss of BH HA represent o one of the mos st costly y unplanned dr rilling events t that can occur for an operator r. Time e spent fishing, , wasted rig tim me and all the a additional cost ts assoc ciated with thi is event signifi ficantly increas se the cost of a well and often results in ex xceeding the approval fo or expen nditure (AFE). . ng drilling is o D Downtime durin only marginall ly better than it i was a few decade es ago. Opera ators estimate es suggest tha at drilli ing efficiency is approxima ately 50% wh hen taking into accou unt all rig act tivities. A typ pical AFE, wh hich includes a detai il listing of in ntangible and t tangible drillin ng costs, show w intan ngible drilling costs can rang ge from 50-75 5% of the tota al cost. Then operato ors typically ea armark an add ditional 10-25% % of th he total AFE to o cover the un nexpected wh hich equates to o millio ons of dollars. C Cost estimates for down ntime on a rig are no ot straig ghtforward as s operators d document the e expenditure es relate ed to drill strin ng failures dif fferently. So t this calculation n

J. Douglas

AADE-11-NTCE-31

found is based on fairly conservative round figures. If 50,000 wells are drilled each year worldwide, ranging in day cost from $45,000 on land to $450,000 for semisubmersible to $1 million on a drillship, the average daily cost could be approximately $250,000 per day. Therefore, based on an estimate of 6 days of downtime for a well, the average cost is $1.5 million per well. Why Taper Threads Dont Fit? The notion that the root cause of most rotary shouldered connection damage and drill string failure is the lack of thread cone interference in connections is not new. In January 1949, American Machinist published Why Taper Threads Dont Fit written by W.E. Burndrett, sales engineer of Pratt & Whitney. In this paper, Burndrett discussed the difficulty that manufacturers experienced in maintaining their standards and controlling the interchangeability of the threaded ends. He pinpointed how a slight lead error of a taper threaded product leads to rapid change in pitch diameter and then small changes in pitch diameter produce large changes in axial travel which leads to variations in gauge standoff. Sixty years ago, Burndrett identified that gauge tolerances ultimately affect gauging results and produce discrepancies. His article advocated for examining each element of the thread with a specific focus on lead, which is a required measurement today. In conclusion, Burndrett stated: No solution can be offered to relieve this problem, except to stress the necessity for trying to reduce errors in individual elements, with particular emphasis on lead, and to exert more vigilance over each phase of manufacture and particularly on gauges in order to reduce the variable to a minimum. This acclaimed article has been referenced for over 60 years and various inspection methods evolved from his analysis. However, the problem of taper threads not fitting still exists and the reason was brought to light by Burndrett long ago small changes in pitch diameter produce large changes in axial travel. Drilling methods have evolved over the years and have become much more severe placing a tremendous amount of stress on the string and more specifically on the connections. Gauging technology has also evolved and its now possible to measure every element over each phase from manufacturing to the field.

class quickly became the industry standard for the most part. Since todays drilling practices are pushing drill strings and particularly their connections to the limits and beyond, the standards are truly being taken beyond their limits. The stresses imposed on the drill string during drilling operations can result in rapid accumulation of fatigue damage and regular inspections are performed to detect mechanical damage and cracking in the rotary shouldered connectors and in the tubes of drill pipe. Normally, rotary shouldered connections undergo visual, dimensional and either magnetic particle or penetrant inspections. RP7G does not have any guidelines for thread inspection nor does it address connection size in reference to fatigue damage and wear. As an inspector examines any element of a drill string, its very important to be cognizant that the acceptable standards being used are arbitrary and not necessarily adequate. Specifications for Threading and Gauging of Rotary Shouldered Connections The API Spec 7-2 1st Edition (released in 2008) is the Specification for Threading and Gauging of Rotary Shouldered Thread Connections. It specifies requirements on rotary shouldered connections for use in petroleum and natural gas industries, including dimensional requirements on threads and thread gauges, stipulations on gauging practice, gauge specifications, as well as instruments and methods for inspection of thread connections. API Spec 7-2 outlines five basic thread attributes that control the size of the new and recut pin and box connectors. Connectors The standoff of the working gauge is Standoff: intended as a method to locate the plane of the pitch diameter in relation to the sealing shoulder of the connection. It is dependent on the other elements of the thread, notably lead and taper. (Figures 1 and 2) Lead: The distance a screw thread advances along a taper cone in one revolution. The increase in pitch diameter of the thread in inches per foot. The distance between the crest and root, normal to the axis of the thread. The form of thread is its profile in an axial plane for a length of one pitch.

Taper:

Inspecting the Drill String The American Petroleum Institute (API) released Recommended Practice RP7G in 1970 to establish an industry wide standard for accepting or rejecting used drill string components. This release established five arbitrary classes of drill pipe, numbered 1 (new) to 5 (junk). Classes 2 to 4 represented graduated levels of pipe displaying signs of wear and tear. Very quickly the industry determined that the best of class (class 2) was not sufficient for many applications, therefore the premium class was added by 1975. Premium

Thread Height: Thread Form:

AADE E-11-NTCE-31

Pr revent Rotary Shouldered Conne ection Failures

lts in larger box connectors s and smaller pin connector rs resul and d doesnt really f fix the problem m. Dam maging Facto ors T The most accu umulating fact tor that leads to drill string g failur re is fatigue, ac ccounting for m more than 40% % of all failures s. Rotar ry shouldered connections t typically fail f from fatigue in n the la ast engaged thr read root of either the box or r pin. Grueling g rig o operations exp pose these th hreads to greater wear and defor rmation than fully engaged d ones. Sin nce most CNC C threa ading lathes produce a -0.0 0005 lead err ror with thread taper r set to nom minal toleranc ce of .015/fo foot, additiona al clear rance is create ed between the e box and pin n threads which h ultim mately leads to o reduced thr read interferen nce. Reduced threa ad interference e allows movem ment and incre eases the stress betw ween load flan nks and resu ults in attritio on. Eventually y, threa ads become wo orn to a point w where there is r reduced contac ct betw ween the pin an nd box threads as shown in Figure 3. As thi is occur rs, the last en ngaged thread d(s) (LET) are e the first 2-3 threa ads near the s shoulder of the pin and the en the first 2-3 threa ads near the cou unter bore of th he box, thus th he total stress in n the c connection is d distributed ove er a smaller are ea at a location n in the e box with less s cross-section nal area. The combination of o these e factors reduc ces the life of f both the pin n and box, and conse equently result ts in premature e connection fa ailures.

Fig gure 1: Plug Gau uge Standoff

Fig gure 2: Ring Gau uge Standoff

Fo or over 92 yea ars, the industry y has relied on n hardened and d groun nd ring and plu ug gauges to inspect i the pitc ch diameter of f rotary y shouldered connections. Its the one feature on a conne ector that has been b left to inte erpretation by the fit of the e ring o or plug gauge e. Basically, the plug gau uge is screwed d hand tight into the e box connecto or. The thread d is within the e permi issible tolerance when the gauging g face of o the working g plug is not more th han -.000 to +.010 + from be eing flush with h the en nd of the shou ulder (Refer to o Figure 1). In I gauging pin n thread ds, the ring gauge g is screw wed hand tight t onto the pin n thread d. The thread is within the e permissible size when the e gaugi ing face of the e working ring g is -.005 to +0.10 of the e nomin nal allowable standoff s (.625 ) (Refer to Fig gure 2). If f the gauge screws on as spec cified, the oper rator presumes s the pr roduct is withi in tolerance, and a when the gauge g does not t screw w on as specified, the operato or presumes th he product will l not fi it. The standard d remedy for a non-fitting gauge g is more e mach hining. Unfor rtunately, this additional machining m only y

Figur re 3: Magnified v view of stab flank worn due to fretting, caused d by in nsufficient threa ad interference. .

V Violent velociti ies, severe vibr rations, whippi ing, and reverse bend ding from stick k-slip and excessive torques a all contribute to o fatigu ue and stress s that lead to o most connec ction damages s. Stick k-slip is the ro otary accelera ation and dece eleration of the drill string which generates con ntinuous torsio onal vibrations s. mentarily at re egular interval ls The bit can stop rotating mom ing the drill st tring to windu up and the torq que to increase e. causi As th he drill tube tw twists and con ntracts, the length of the dril ll string g becomes sho orter. Subsequ uently the stuc ck drill bit lift ts off th he bottom of th he hole allowi ing all the store ed energy from m the d drill string to produce rever rse rotary spee eds up to three times s the drilling R RPM. mmon Forms s of Damage Com Threa ad damage is noted as t the most com mmon problem m encou untered with the use of d drill pipe. One e of the main n chara acteristics of thread dama age is that i it will spread d

J. Douglas

AAD DE-11-NTCE-31

throu ughout the dril ll string as on ne damaged th hread connects s with another. Thre ee of the most t common forms of damage e found d in connector rs are shoulde er galling, box x swelling and d stress s cracking. Ea ach are the res sult of a stress s concentration n encou untered at the last thread ma ade up in a con nnection. This s stress sed concentrat tion coupled with w wobbling g and cycling g stress s of the connec ction due to im mproper thread make up from m lack o of interference e results in a hi igh incidence of o pin and box x failur res. y design, there By e is an inheren nt lack of threa ad interference e due to o the allowable e tolerances fo or thread taper in pin and box x conne ectors. This tolerance t allow ws the box to o axially pivot t again nst the pin shou ulder and the first f thread from m the pin nose e on the opposite side e which produc ces an oversize ed oval shaped d box o opening (Refer to Figure 4 an nd 5).

shoul lders causes ex xtremely high contact pressu ure that deform ms both faces.

Fig ure 6: Diagram of a connection n bending. Pin t thread fretting, ss cracks and b bell mouthing re esult. box stres

Figure e 4: Box taper with w a tolerance in the minus di irection and the e lack of th hread interference is circled in red.

F Figure 7: Diagra am of bell mout thing of the box x connector. Result of f undersized pin n and oversized d box.

Figu ure 5: Pin taper with w a tolerance e in the plus direction and the lack of th hread interference is circled in red.

W When a conne ection starts pi ivoting or roc cking back and d forth h at the axis cr reated by the c clearance betw ween the thread stab f flanks, the 30 angled flanks begin to wear r and look like a radiu us. This damag ge is induced u under load and in the presence of rep peated relative e surface motio on, better know wn as fretting as sh hown in Figure es 6 and 7. The e fundamental way to preven nt fretti ing is to amen nd designs which allow no r relative motion n (axia al bending) of t the surfaces tha at are in contac ct. Desi ign D Designers today y draw RSC co onnectors utiliz zing various 3D D mode eling programs; the design is then analyz zed using finite elem ment analysis ( (FEA) softwar re to evaluate stresses unde er speci ific loads. Th he major flaw with this proc cess is that the total allowable tole erances for thr reads (see Tab ble 1) produced d

A At this point, because b the box is out of shape and the e weak ker of the two members, the e pin can be sc crewed farther r into t the box causin ng the box opening to flare like a bell. As s this h happens, the bo ox and pin sho oulders are no longer parallel l to one another and the decrease in surface cont tact area of the e

AADE E-11-NTCE-31

Pr revent Rotary Shouldered Conne ection Failures

durin ng the manufa acturing process are not en ntered into the e equat tion.
Table e 1:

Current AP PI Specification 7-2 Thread Attribute Tole erances of RSC C Connectors

Thread Attributes Sta andoff Gau uge to Sho oulder .005" -. +.010" .000" . +.010" Thread Lead Thread Taper Half f Angle Threa ad Form m 4 30 or 45 , deg g 0.75 5 30 or 45 4 , deg g 0.75 5 Thread Height +.001" -.003 +.001" -.003 +.001" -.003 +.001" -.003

Pin Thread Attributes

+/.0015/in

0/ +.030

Box Thread Attributes

+/.0015/in

0/ -.030

In n other words, analysis a is base ed on models with w no lead, pitch diameter, taper, or thread form errors, which w explains s why connectors wi ith slight threa ad errors create potential for r failur re. Only wh hen models are a drawn with w maximum m allow wable thread at ttribute toleran nces for both connectors c will l designers see the total effects s of reverse bending and d whipp ping. Analysis Sever ral FEA mode els were constructed and ev valuated for a NC50 0 tool joint with w various attribute a tolerances. Design n param meters are show wn in Figure 8.

re 9: Interferenc ce of the threads on assembly using 9,167 ft-lb b Figur torque - resolved but no o shoulder cont tact.

Axia al Stress Axia al stress is a t tension or com mpression stres ss created in a struc ctural member by the applic cation of a le engthwise axia al load. . The followi ing models il llustrate the hot spots or o poin nts of highest s stress of the connection expe erienced during g make eup. In Figure 10, 9,167 ft-lb bs was resolved d for zero pitch h diam meter and .001 1 lead with no shoulder loads s. In Figure 11 1, 30,00 00 ft-lbs was r resolved to allo ow box should der contact with h a con ntact pressure of 20,000 ps si. The LET already show ws signs s of stress bu ut the threads are in interfe erence and the conn nection should p perform as inte ended.

Figu ure 8: FEA Desig gn Parameters

Fo or the purpose e of this study, Youngs mod dulus 30e6 psi, , Poiso ons ratio 0.3 an nd a friction fac ctor of 0.1 were applied. The e follow wing models illustrate the ho ot spots or points of highest t stress s that the conne ection experien nced. Figure 9 illustrates the e interf ference of the threads on as ssembly with 9,167 9 ft-lbs of f torqu ue.

Fig gure 10: Box Pit tch Diameter +.000; Pin Pitch D Diameter -.000 /Lead Errors B ox +.001 and Pi in -.001; 9,167 ft t-lb torque; No Shoulder Loads.

J. Douglas

AAD DE-11-NTCE-31

F Figure 11: Box Pitch P Diameter +.000; + Pin Pitch Diameter .00 00/Lead Errors Box B +.001 and Pin P -.001; 30,000 0 ft-lb torque

inally in Figu ures 12 & 13 3, using the pitch p diameter r Fi tolera ance of .012 2 which is th he maximum allowed pitch h diame eter tolerance using ring and a plug gaug ges, the stress s exper rienced by the entire connect tion is obvious s, especially in n the box. By examining the exploded view in Figure 13, the e lack o of thread interf ference is evid dent as well as the number of f critca al hot spots. Essentially, th his box and pi in would have e passe ed inspection using ring and plug gau uges but this s conne ection may not be able to en ndure what is required down n hole.

Figur re 13: Closer Lo ook at Figure 12 2.

In n summary, the e connections shown in Figures 12 & 13 are exam mples of oversiz zed boxes and undersized pin ns which would d using the cur have been approved rrent API sanc ctioned ring and d plug gauging meth hods. Howev ver, most imp portantly, these conn nections would d NOT have been approve ed if the pitch h diam meters were me easured with a pitch diamet ter gauge using g tolera ances of .002 2 and .004. ution Solu T Typically rota ary shouldered connection n failures and dama ages are from the lack of thr read cone inte erference which h can be directly re elated to the pitch diamete er size and it ts accep ptable level o of tolerance. As shown in n figures 9-12 2, curre ently allowable e pitch diamet ter tolerances can drastically y affec ct the integrity of a connectio on. Simply by measuring and contr rolling the pi itch diameter, the number of threads in n interf ference increa ase when the connections are made up p. Being g able to a accurately me easure the pi itch diameters s, indep pendent of the e other thread elements, allo ows the size of o the c connection to b be regulated du uring manufact turing and more impo ortantly once t the connector has been use ed. Without a mean ns of measurin ng the pitch d diameter indep pendently, each h conn nectors pitch d diameter can be affected from m .016 - .019 by th he cumulative e errors in lead, taper and threa ad form. These error rs lead to oversized box co onnectors and undersized pin n conn nectors. metrically a .00 01 Lead Error generates a .0 0034 diamete er Geom chang ge as illustrat ted in Figure 14. Lead Er rror / tan30 = Radia al change (.00 01/tan30=.0017) x 2 = .0034 4 on Diamete er for ev very inch of th hread.

Fig gure 12: Box Pitc ch Diameter +.0 012; Pin Pitch Diameter -.012 / /Lead Errors Bo ox +.001 and Pin n -.001; 30,000 ft t-lb torque

AADE E-11-NTCE-31

Pr revent Rotary Shouldered Conne ection Failures

Figure 14: Effects of lead er rror on pitch dia ameter.

Le ead error has s a particular rly drastic ef ffect on pitch h diame eter. Threads that t have allow wable lead or other attribute e errors s handicap perfectly ground d ring and plu ug gauges and d make e the gauges standoff prematu urely. As a re esult, when the e gauge e standoff is too much, the machinist t immediately y remov ves more mate erial which is not necessarily y the right fix. Howe ever, by sizing g the connector correctly wit th the addition n of a p pitch diameter r measurement t, the expenses s of reworking g conne ectors later can n be drastically y reduced. A All of the curre ent API specif fication thread d attributes are e stand d-alone measur rements and have h their own n tolerance, all l excep pt pitch diamet ter. Pitch diam meters historic cally have been n contro olled by ring and plug gaug ges and were never given a tolera ance or gauging g method. Rin ng and plug gau uges have been n used to size the co onnectors func ctional pitch diameters d at a design gauge plan ne located 5/8 from the pin and box x shoul lders. The gaug ge plane is con ntrolled axially y by tolerance. Depending on conn nector size, the pin diameter sized s by a ring g gauge e occurs betwe een 1.257 to 1.375 from the pin torque e shoul lder and in a box connector r between 1.125 to 1.500 from the box face. Figure 14 displays the effec ct lead error or r s on pitch diam meter. axial movement has
Table e 2:

diam metrical design n size of the e connection without being g affec cted by other th hread attributes s. B By isolating the connectors pitch diameter r from the ring g and plug gauge standoff, pitch diameter gau uges provide a check k and balance that ensures th hat more thread d flanks go into interf ference when t the connection ns are made up to the designed torqu ues. By design n, threads shou uld be in interf ference prior to o or at the same time the shoulders make contact. W When pitch d diameter tole erances are c controlled, the conn nections taper red thread con nes and should ders will be in n interf ference. Torq ques and bendi ing forces will l be distributed more e evenly over r the entire co onnection redu ucing shoulde er gallin ng, box swelli ing and possib bilities of stres ss cracks. AP PI Spec c. 7-2 connecti ion designs can n experience h higher levels of o perfo ormance and lo ongevity with t the addition of f pitch diamete er meas surement. C Currently, ther re is no spe ecification to control pitch h diam meters on use ed connection ns even thou ugh boxes are comm monly measur red at +.018 to +.028. U Used connector rs are e easily measure ed in the field for box swelli ing and loss of o threa ad engagement t. Better docu umentation, ins spector training g and 3 3rd party verif fication of thre ead attribute to olerances at the manu ufacturing level will en nsure product ts are being g manu ufactured to AP PI tolerances. Con clusions T The vibrational l and impact loads experien nced by rotary y shoul ldered connec ctions result i in various fo orms of thread dama age and failure es including sh houlder galling g, box swelling g, threa ad fretting, stre ess cracking, an nd ultimately, twist offs. It is i wide ely accepted th hat stress and fatigue of rot tary shouldered conn nections due to o frictional co ontact between n box and pin n threa ads are comple ex problems. U Unfortunately, the complexity y of fa atigue and the e designers u unawareness to o many of the facto ors affecting va arious elements of the connec ction is anothe er chall lenge. Several l approaches t to alleviate fati igue have been n avail lable for years s, and many a applications ha ave been tried d; howe ever very few h have achieved widespread su uccess. B Beginning with h its first st tandards in 1924, API ha as main ntained the API I Spec.7-2, Specification for r Threading and Gaug ging of Rotary y Shouldered T Thread Connect tions. For ove er ears, the same 87 ye e tired method d of sizing con nnections pitch h diam meters has been n with ring and plug gauges. F Forward thinkin ng operators a and inspectors are now taking g advan ntage of mor re up to dat te gauging te echnology and perfo orming in-dept th inspections of their pipe. Now operator rs are a able to inspec ct both new a and used pipe e threads using g preci ision pitch dia ameter gauges that allow them to pinpoin nt early y signs of stress, perform pr reventative ma aintenance, and exten nd the life of th heir drill string g connections. W When the pitc ch diameter t tolerances are e more tightly y contr rolled, the c connections tapered threa ad cones and shoul lders will be i in interference e. Therefore d drilling torques and b bending forces s will be more e evenly distri ibuted over the entire e connection reducing dam mage. By ad dding the pitch h diam meter measurem ment to API S Spec 7-2 curr rent connection n

Necessary y RSC Product Attribute A Tolera ances

Thre ead Attrib bute Thread Axis to Shoulder Standoff Gauge G to Shoulder S -.005" .002" +.010" .000" .002" +.010" Thread Lead Thread Taper Half Angle Thread Form 30 or 45 0.75 30 or 45 0.75 Thread Height + .001" - .003 + .001" - .003 + .001" - .003 +.001" - .003 Pitch Diameter Tolerance 2 TPF - .001 + .002 .000 + .002

Pi in Thre ead Attrib butes Bo ox Thre ead Attrib butes

+/.0015/in

0/ +.030

+/.0015/in

0/ -.030

he difference between the e API stando off and pitch h Th diame eter size on pin n and box con nnections is the e result of lead d errors s. Lead errors s affect ring an nd plug gauge es, but have no o effect t on pitch diam meter gauges. The same sce enario explains s why ANSI has calc culations for pitch p diameter tolerances for r UN, A Acme and Stub b Acme threads because they y recognize that t pitch diameter is the t only thread attribute tha at controls the e

J. Douglas

AADE-11-NTCE-31

design, connections can perform more efficiently and last longer. Amazingly after new connectors have been coated or used, no other means or method of measurement is currently required to verify the reliability of thread elements. Adding precise pitch diameter measurements to API RP7G will enable operators to detect fatigue, box swelling and loss of thread engagement in the field and on the rig floor earlier and usually BEFORE stress cracks occur. Finally, a response is offered to Burndretts final conclusion that No solution can be offered Sixty two years later, a solution to relieve the problem of errors in individual elements with particular emphasis on the lead as well as the equally important pitch diameter is available. The application of precision gauges measuring lead, taper, form, height, and pitch diameter can and will minimize error and prevent rotary shouldered connection failures.

AADE E-11-NTCE-31

Pr revent Rotary Shouldered Conne ection Failures

Case e Study 1: Strat taLoc Implem ments New In nspection Pr rocedures R Rick Nichols, BHA B Designer r with StrataLo oc Technology y was s searching for a way to hold closer toleranc ce limits on all l the th hread attributes s of their RSC connections. Because all of f Strata aLocs high tensile straight ACME A threads are currently y being g measured to o ANSI Syste em 22 using Gagemakers s functi ional roll and d pitch diamet ter gauges, Ri ick decided to o inquir re about a sim milar gauge fo or tapered thre eads. He then n learne ed of Gagemak kers Joint Stre ength System (JSS) for rotary y shoul ldered connect tion inspection ns. He immed diately ordered d what he needed and a proceeded d to measure all his NC70 0 conne ections as a tria al. He anticipa ated that measuring the pitch h diame eters of the NC70 connections would ensure more e interf ference in the threads t prior to t shouldering and resist any y thread d damage from m the stress exp perienced down n hole. H He inspected 15 5 sets of NC7 70 connections s with the JSS S and th he readings we ere recorded as s found in Table 3. The other r conne ections in the string, s a 6 5/8 API Reg. box x by 7 5/8 API I Reg. pin bit sub, a 7 5/8 API Reg g. box crossove er (drill bit Sub b to BH HA) and the NC C56 crossover were produced at an outside e vendo or and measur red using only y the minimum m API required d attributes which inc cluded ring and d plug gauges. In n December 20 009, the BHA tool t was run do own hole for 8 days in a well sout th of Oklahom ma City to a TD T of 12,000. Then, , the rental to ool was sent back to Hous ston, TX. On n Decem mber 23, 2009, 2 a re-inspection wa as performed d incorp porating the JS SS inspection into their norm mal inspection n proce ess. Because the t pitch diam meter gauge can n also be used d on us sed pipe, these measurements were captured d again.
Table e 3:

Figur re 15: Damaged d threads declar red by Lewis Ins spection Group p.

T They passed th he 15 NC70 c connections as s there was no exces ssive damage. Only two box xes showed sl light galling on n secon nd thread from m stabbing dam mage. This wa as field dressed and r ready to go. T These results th hrilled Nichols s and StrataLoc Tech hnology now r requires all tap pered threaded d connectors be meas sured using the e JSS which pr rovides precise e measurement ts for le ead, taper, height and most t importantly p pitch diameter r. Since e implementin ng this new in nspection in th heir procedures s, Strata aLoc has seen n an 85% imp provement in p performance of o their connections an nd has had few wer failures in t the field.

Th he third party inspection group, Lew wis Inspection n Comp pany, rejected the 6 5/8 Regu ular box X 7 5/8 5 Regular pin n bit su ub, 7 5/8 Regular box API cross over (d drill bit Sub to o BHA) and the NC56 cross over to o the drill string g for excessive e d damage per current API RP7G G inspection n thread specif fications. The e damaged part ts, shown in Fi igure 15 & 16, , were from the set of o connections only inspecte ed to minimum m requirements which w included d ring and plu ug gauges and d API r not JS SS.

re 16: 7 5/8 API Regular display yed evidence of f thread fretting g. Figur After inspe ection, it was de etermined the pi in pitch diameter was s .021.

10

J. Douglas

AAD DE-11-NTCE-31

Case e Study 2: Tejas s Tubular Im mprove Qualit ty and Perfor rmance Er ric Otten, Dr rill Pipe Operations Manager for Tejas s Tubu ular of Housto on, TX which h is the large est supplier of f afterm market Horizon ntal Directiona al Drilling tube es (HDD), had d been searching for ways of contr rolling tighter pitch diameter r tolera ances on their HDD product t line to increa ase connection n performances in ben nding and high h torque applica ations. A After closely reviewing r thei ir internal pro ocesses, Otten n found d that standof ff varied grea atly and that he needed to o scruti inize the insp pection method d for each th hread attribute. Some e of his other discoveries d included: Lead errors in the CNC machined product p threads s in nterfered with their t certified ring r and plug gauges ground d le ead and taper causing premature standoff f to the datum m to orque shoulders s. The incr reased amount t of gauge st tandoff was a ca alculated amou unt over the len ngth of the con nnections every y tim me. Every .0005/in. lead error e resulted in .002 more e m material being removed from the t connection ns design pitch h di iameter allowa able limits. Increased d material rem moval on the pitch p diameters s re educed all the hoop h stresses in n the connectio on, which were e ne eeded to resist t over torquein ng and bending g forces during g di irectional drilli ing. A After completin ng his internal l investigation n, Otten began n his search and fou und the Joint Strength Sys stem (JSS) by y Gagemaker which focused f on accuracy and prec cision of rotary y shoul ldered connect tions. JSS pr rovided a met thod to check k thread d cone taper, thread t lead, thr read height, thr read form, and d pitch diameter at the pitch plane which w are all th he features that t make e a connector perform prop perly. He imm mediately was s intrig gued by the adv vantages of th he JSS pitch di iameter gauges s for m monitoring an nd holding pitch p diamete ers to tighter r tolera ances as well as a verifying th he repeatability y of their CNC C lathes s on the shop floor. f Otten im mmediately im mplemented the e JSS o on their HDD product p line. Si ince RSC pitch diameters ha ave never been n measured or r docum mented, hundre eds of various sizes were test ted. All thread d attributes were measured m inclu uding lead, standoff, pitch h diame eter, taper and thread form. Table 4 notes a sampling of f the va alues found.
Table 4

A After great s success with HDD produ ucts, he then n inves stigated their API RSC con nnections to see if the same effec cts on their ring g and plug gau uges were affec cting the size on n their pin and box c connectors. An nd, as expected, he found the same e problems, so o he immediat tely put the J JSS into action n addin ng the pitch d diameter measurement to t their API RSC C inspe ection proced dures for do ocumentation, tracking and mark king as illustra ated in Figure e 17. Since this procedure chang ge, RSC produ ucts have also o seen a drama atic increase in n perfo ormance level.

Figure 17: Teja as stamps each h connection wit th JSS INSP whi ich shows that t they have not o only been inspec cted to current A API Specification ns, but the prec cise pitch diame eter has been m measured and d documented.

Sampling g of Box and Pin n Measured Values

Lead -.001 -.0005 .000 +.0005 +.001 -.002/.000/+.002 St tandoff from Nominal .000 .000 .000 .000 .000 .000 Pitc ch Diameter .006/.012 +.003/.006 +.002/.004 +.003/.006 +.006/.012 +.01 18 Taper .1666/.250 .1666/.250 .1666/.250 .1666/.250 .1666/.250 .1666/.250 Thread Form

AADE-11-NTCE-31

Prevent Rotary Shouldered Connection Failures

11

Acknowledgements
The author gratefully acknowledges the following people for the help and contributions to this paper: Scott Benzie, Mohawk Energy, Houston, TX Rick Nichols, Strata Loc Technology, Houston, TX Eric Otten, Tejas Tubular/Premier Drill Pipe, Houston, TX John Wolfe III, Gagemaker LP, Pasadena, TX

References
1. 2. Burndrett, W.E.: Why Taper Threads Dont Fit, American Machinist, McGraw-Hill Publishing, 1949. Van Der Wissel, H.T.: Box-and Pin Type Threaded Joint Having Different Pitches and Pitch Diameters, United States Patent # 307,9181, Published Feb. 26, 1963. Farr, A.P., Hughes Tool Co.: Torque Requirements For Rotary Shouldered Connections And Selection of Connections For Drill Collars, paper ASME Petroleum Division, No. 57-PET-19, presented at the ASME Pertoleum Mechanical Engineering Conference, Tulsa, Oklahoma, 22-25, September 1957 Reed Tubular Tool Joint United States Patent # 4549754 Published 1985 ASM Handbook, the American Society for Metals, Vol.13 "Corrosion", ASM International, 1987. Collins, T. and Vaghi, F.: Analysis of the Fatigue Resistance of Rotary Shouldered Connections, paper 74564-MS presented at IADC/SPE Drilling Conference, Dallas, TX, 26-28 February 2002. Hamid, K. and Shibu, J.: Drilling String Inspection-The Dilema Between: 1) Oil Operating Companies, 2) Drilling Contractors and 3) Tubular Inspection Companies, paper preented at Middle East Nondestructive Testing Conference & Exhibition, Bahrain, Manama, 27-30 November 2005. Berry, C.: Drilling Failure Costs Quickly Add Up, JPT, August 2009. York, P., Pritchard, D, Dodson, J., Dodson, T, Rosenberg, S and Utama, B: Eliminating Non-Productive Time Associates with Drilling Trouble Zones, paper presented at Offshore Technology Conference, Houston, Texas, 4-7 May 2009. MacDonald, K.A., Deans, W.F.: Stress Analysis of Drillstring Threaded Connections Using The Finite Element Method, Engineering Failure Analysis, Vol. 2, No. 1 pp. 1-30, Aberdeen, U.K. 1995. Kessler, F., Muradov, A., Smith, J.E.: Fatigue Resistent Disposable Drill Pipe for Short Radius Applications. Leine, R.I., Van Campen, D.H., Keultjes, W.J.G.: Stick-slip Whirl Interaction in Drillstring Dynamics, Journal of Vibration and Acoustics, Vol. 124 pp. 209-220, ASME, 2002. Baragetti, S., Baryshnikov, A.: Rotary Shouldered Thread Connections: Working Limits Under Combined Static Loading, Journal of Mechanical Design, Vol 123, Issue 3, 456 (8 pages), Italy, September 2001. Tollefsen, E., Weber, A., Crowell, D. and Schlumberger, Z.: Evaluating the True Cost of Losing a Bottomhole Assembly, paper 112704 presented at IADC/SPE Drilling Conference, Orlando, FL, 4-6 March 2008. Baryshnikov, A., Anatoly, Ligrone, A., Ferrara, P.: Optimization of Rotary Shouldered Connection Reliability and Failure Analysis, paper 27535-MS presented at SPE/IADC Drilling Conference, Dallas, TX, 15-18 February, 1994. Kaiser, M.: A Survey of Drilling Cost and Complexity Estimation Models, International Journal of Petroleum Science and Technology, Vol. 1, Number 1, pp. 1-22, 2007.

3.

4. 5. 6.

7.

8. 9.

10.

11. 12.

13.

14.

15.

16.