field testing of casing design factors

Field Testing of Casing-string Design Factorst J. E. SAYE* AND T. W. G. RICHARDSONG

th l s method, i t was reasonable t o belleve that the deslgn factors were conservative. Therefore, In May 1952, a field-testing program was undertaken In the Elk C ~ t y Fle ld , Oklahoma, t o a s c e r t a ~ n whether design factors for tension and collapse could be lowered safely. Th l s paper presents up-to-date re- s u l t s of t h i s test lng program. Tenslon design factors were progressively lowered from 1.60 to 1.40 In increments of 0.05 wlthout Incurring tens l le failures. In the collapse t e s t s , uncemented cas ing was collapsed a t design factors of 0.95, 0.85, and 0.80.

ABSTRACT

ing design factors as standard for the Elk C ~ t y F ~ e l d : tension d e s ~ g n factor of 1.40, collapse design factor of 0.85 for cemented caslng, collapse d e s ~ g n factor of 1.00 for uncemented casing; and the burst design factor of 1.33 remaining unchanged. Use of these deslgn factors has, however, been re- s tr lcted t o t h ~ s one field. As soon a s the field-testing program has been completed and carefully ana- lyzed, a revised caslng deslgn method and set of d e s ~ g n factors w ~ l l be prepared for unrestricted use.

For about 10 years Shell Orl Company h a s designed combinat~on cas lng s t r ~ n g s usmg an engineered standard method and design sa fe ty factors of 1.60 ~n tension, 1 .OO m collapse, and 1.33 ~n burst. In the absence of any cas ing fai lures attributed to

INTRODUCTlON

However, cemented casing successfully withstood col lapse d e s ~ g n factors as low as 0.50, ~f design factors are calculated in exactly the same manner a s for uncemented casing. A s a result of the t e s t da ta presented hereln. Shell h a s a d o ~ t e d the follow-

Through much of i t s recent h ~ s t o r y the 011 lndus- try h a s been plagued with shor tages of tubular goods and government control of c r i t i ca l materials. T h ~ s h a s forced the industry t o depart from ~ t s proved methods of caslng-str ing design and h a s led e ~ t h e r t o arbitrary r e d u c t ~ o n of d e s ~ g n fac tors or wasteful use of h e a v ~ e r or b e t t e q r a d e pipe than that ordin- arlly r e q u ~ r e d . At the same tlnle the never-ending s e a r c h for new petroleum reserves h a s resul ted in drilllng t o Increasingly greater dep ths which re- quires more exact ing des ign of well cas ing . Also, a n ana lys i s of the c o s t of wel l s drilled In the mid- continent a rea by Shel l Oil Conipany durlng 1953 s h o w s that well cas lng represen ts 27.5 percent of the drllllng cos t . All of t h e s e fac tors s e r v e to em- phas ize the f a c t that c a s ~ n g - s t r ~ n g design IS a n en- p e e r i n g problem of great lmportance.

In order t o cope with th i s problem the 011 Industry must have h igh-gade tubular goods with well-estab- l ished s trengths, a rel iable method of c a s ~ n g - s t r i n g des ign , and rea l i s t i c des ign fac tors . Considerable work h a s already been done In developing better * Shell 011 Company, Oklahoma C ~ t y . hell 011 Company. Tulsa. f Presented by J. E. Saye at the sprlng meetlng of the Mld-

Con t~nen t Dlstrlct, Dlvlslon of Production, Oklahoma C~ty, March 1954.

mater ials and in statistically e s t a b l i s h ~ n g s t reng ths of tubular ,goods from t e s t da ta . Also, s e v e r a l reli- ab le design methods have been presented in the literature. T h e des ign factors In use , however, have been a r b ~ t r a r y and vary c o n s ~ d e r a b l ~ among the op- erat ing oil companies. Very l ~ t t l e h a s been done ex- perlmentally t o e s t a b l ~ s h t h e s e d e s ~ g n factors .

Durlng the ea r ly part of World Rar 11, Shel l 011 Company adopted a s tandard method of c a s i n g - s t r ~ n g des ign using the following design factors: 1.60 for tension, 1.00 for co l lapse , and 1.33 for burst. S ~ n c e adoptlng t h i s method, a great number of cas ing s t r ings have been run and no fai lures have been attributed t o the d e s ~ g n method or the d e s g n factors

-

used. T h i s absence of failure w a s good reason for b e l i e v ~ n g that the design fac tors i n use were con- se rva t ive and might be reduced. Therefore, In May of 1952, Shel l Oil Conipany undertook a field-test program t o determine whether or not co l lapse and t e n s ~ o n design fac tors could be lowered. T e s t s were conducted In field w e l l s because field experience i s the only yardst ick which h a s a n y r e a l meaning.

T h e purpose of t h i s paper 1s t o present the resu l t s which have been obtalned t o da te with . this field t e s t l n g of casing-string des ign factors . A s the math- enlat ical and phys ica l concepts of c a s i n g - s t r ~ n g de- s i g n have been covered in previous papers , the

24 J . E . SAYE AND T. W. G . RICHARDSON I

writers do not plan to d iscuss them. The casing- strength values used in this report are those published In API Bu15C2, Fifth Edition, January 1952. Also, this paper must necessa r~ ly be considered a progress report because the tes t program has not yet been completed and tes t results have not been used to establish design factors for unrestricted use. Throughout this paper the writers will refer to "design factors" rather than "safety factors" a s true safety factors are not known. It should again be pointed out that only tension and collapse design factors have been field tested in wells because it was inlpractical to conduct burst tests . Cas~ng-string Lles~gn Rlethods

The principles of casing-string design were se t down a s early a s 1921; and, in 1941, theoretical and practical investigation on the part of both the pipe industry and the oil Industry culminated in pub- lication by the American Petroleum Institute of nilninlum strengths for casing in collapse, tension, and burst. Today, engineered casing-string design is the rule rather than the exception and a number of acceptable design methods have been published in the literature, but there seems to be little uni- formity in the design factors used. j4 recent survey conducted by an 011 conlpany operating in the m ~ d - continent area showed that the following design factors are used:

Collapse Design

Company* Factor A 1.125 l3 1.125 C 1.125 D 1.17 E 1 .oo F' 1.125 G 1.125

Tension D e s ~ g n Factor

1.80 1.80 1.75 1.75 1.60 1.60 1.80

Burst Design Factor 1 .oo 1.10 1.10 1.75 1.33 1 .oo 1.10

In general, these companies base their design upon API minimum y~eld-strength values but limited use is a l so made of ultimate strength values. Some companies use actual mud weight in design, whereas others use the weight of sa l t water. In general, bouyancy IS neglected but usually bi-axial s t r e s s curves are used to determine the reduction of collapse resistance caused by tensile load.

For a number of years Shell Oil Company has fol- lowed a standard method for designing tapered cas-

*Not necessarrly In the order glven, these comp-es were Shell 011 Company, Stanollnd 011 and G a s Co., Phllllps Pet- ro leum Co., Gulf 011 Corp., Deep R o c k 011 Corp., The Carter 011 Co., and Sunray 011 Corp.

ing strings using niininiuni strength values published by the API and design factors of 1.60 in tension, 1.00 in collapse, and 1.33 in burst. In designing for collapse, the full mud-column pressure is presumed to be applied on casing empty of fluid (~ . e . , an internal pressure of 0 psig is assunled), bouyancy is neglected, but bi-axial s t r e s s curves are used to determine the ellect of t e n s ~ l e load on collapse resistance. In designing for tension the free-hanging weight, calculated in air, is used, bouyancy is neglected, and the t e n s ~ o n design factor is based on

-

API niin~n~unl jolnt pull-out strength. Burst strength 1s usually not critical but calculations are nrade based on the niaximum antrcipated internal pressure, and external pressure IS presumed to correspond to that of a column of salt water. This design method was used to prepare the normal 5 '/,-in. casing-string design chart, Fig. 1, for the Elk City Field. Prior to the field testing of cas~ng-string design factors, thls was our standard design for the field.

Throughout this paper, cas~ng-string design factors will be calculated a s out l~ned in the foregome, d i s c u s s ~ o n of the Shell design method.

DEPTH IN FEET 9500 10000 10500 11000 11500

Mud We~ght: 9.4 Ib per gal Des~gn Factor, Tens~on: 1.60 Des~gn Factor, Collapse: 1 .OO D e s ~ g n Factor, Burst: 1.33 at 4,000 p s ~

wellhead pressure Fig. 1-Normal 5'4-in. Cosing-string Design Chart

E l k C i t y Field, Beckham County, Oklahoma

F I E L D TESTING O F CASING-STRING DESIGN FACTORS 25

F i e l d Conditions Afiecting Tests T h e E l k C i t y F i e l d in Beckham and Washita

Count ies of Oklahoma w a s se lec ted as the most su i tab le place for conducting t e s t s 'bf co l lapse and t e n s ~ o n design factors because the depth of wel l s o l e r e d considerable flexibility in casing-str ing de- s i g n and t e s t s could be conducted with the l e a s t

-

r i s k of losing a well. A maximum of wel l s a f e t y w a s a s s u r e d because tenslon t e s t s could be conducted s o that the t e s t sec t lon w a s Inside the 3,100-ft in- termediate or "salt-string" cas ing whlch is needed In th i s field. Also, co l lapse t e s t s could be s a f e l y conducted without d e e p e n ~ n ~ wel l s because wel l s frequently encounter severa l hundred f e e t of non- c o n ~ m e r c ~ a l s t r a t a a s a resu l t of t es t ing a lower zone, not productive In a l l par ts of the field. With the well-trained field supervisors and se rv ice per- sonne l avai lable , i t w a s possible t o conduct t h e s e t e s t s with minimum r i s k because , if f a ~ l u r e occurred during a t e n s ~ o n t e s t , the s t r ing would be parted In- s i d e the i n t e r m e d ~ a t e cas ing and repalr would be rel- a t ively simple and inexpensive. In the even t of a co l lapse failure In the bottom of the s t r ing, the hole l o s t would have no conlmercial value and the cas ing could st111 be cemented through perforations. -

- .

A s a l l of the t e s t s reported herein were conduct- e d In the E l k Ci ty F ~ e l d a general d i s c u s s i o n of well condit ions follows. After cementing a n i n t e r m e d ~ a t e cas ing s t r ing a t about 3,100 f t i n a n E l k C ~ t y F i e l d well, a 7'4-in. hole is drilled t o a depth which aver- a g e s 10,500 ft. A 9.4 lb per gal mud i s used, but bottom-hole pressure general ly corresponds to that of a column of-fresh water. T h u s a t 10,500 f t there is a mud over-balance pressure of 578 p s i resulting from the diflerential between 5,124 PSI mud-column pressure and 4,546 p s i formation pressure. Forma- tion bottom-hole temperatures range from 168 t o 182 F., and the S1/,-in. cas ing s t r lng is d e s ~ ~ n e d for a maximum wellhead pressure of 4,000 psi . When the 5 '/,-in. oil-string cas lng i s run, it 1s filled with mud, the lower 1,500 t o 2,000 f t of the s t r ing is equipped with central izers and sc ra tchers , and the pipe i s reciprocated through 25-ft s t rokes while it is being cemented. T h i s r e c ~ p r o c a t i o n imposes a n addition- a l tensi le load which averages 30,000 lb. T h i s load i s not considered when the c a s i n g s t r ing i s designed because it is opposed by bouyancy which i s a l s o neglected. On the l a s t downstroke of rec~proca t ion , t h e cas ing is landed and, if the float holds, cas ing s l i p s a re s e t immediately, i.e., cas ing i s not s lack- e d 05 after the cement h a s s e t . In the E l k Ci ty F i e l d

uncontaminated cenient tops a t 8,500 f t a r e common and friiquently a r e as high a s 8,000 ft. F le ld - tes t Procedure

Insofar a s possible the field t e s t s of casing-str ing des lgn factors were handled as though the cas ing were the same a s for a routine field well. No spe- c ~ a l instructions were Issued t o field personnel be- yond those necessary t o get c o n ~ p l e t e d a t a on the tes t . In performing tension t e s t s reciprocation and a l l a l l l ed o p e r a t ~ o n s were conducted In the usual manner, making no al lowance for the reduced design factor. In fact , field personnel were del iberately a s k e d not t o take any s p e c i a l ca re in the running and cementing of the tension t e s t strings. On col- l a p s e t e s t s the only departure from customary procedure w a s the running of the co l lapse t e s t s e c t i o n on t h e bottoni of the string. Tens ion and co l lapse t e s t s were never conducted in the same well, however , and Moody Engineering Co. inspect ion reports were obtained for a l l cas ing used in the t e s t s e c - t ~ o n s .

T e n s i o n Tests In conducting t e s t s of tension des ign factors ,

c a s l n g s t r ings were designed and run in the usua l manner. T e n s i o n des ign factors , ca lcu la ted on the b a s s of the cas ing weight in a l r and API minimum joint pull-out strength, were progressively lowered in successive cas ing s t r ings in accordance with the following schedule:

T e s t Number T e s t Design E'actor 1 and 2 1.55 3 and 4 1 .50

5, 6 , and 7 1.45 8 1.40

In the event of t ens i le failure, repair would have been faci l i ta ted a s that portion of the c a s i n g s tr ing being tes ted w a s a lways inside the s a l t s t r ing.

All e ight of the t e s t s l is ted were conducted without failure and the r e s u l t s have been tabulated in T a b l e 1. Fig . 2 i l lus t ra tes the comparison of a 1 .60 and a 1.40 des ign factor for tension i n a c a s i n g s t r ing for a well 11,000 ft deep. Fol lowing the s u c - c e s s f u l completion of these t e s t s , a tension des ign factor of 1 .40 w a s accep ted a s s tandard for the E l k Ci ty F i e l d and, t o da te , 4 0 oi l s t r ings of t h l s de- s i g n have been success fu l ly run and cemented a t dep ths from 10,000 t o 11,000 f t .

It is interest ing t o note from T a b l e 1 that , if the bouyancy of the mud and the reciprocation load a re considered, tension des ign fac tors for t h e s e t e s t s

26 J. E. SAYE AND T. W. G. RICHARDSON

Table 1 Cas~ng Tension Tests

Elk C ~ t y F ~ e l d , Beckham and Wash~ta Counties, Oklahoma T e s t No. 1

R. E. Johnson

"B" No. 1 7-28-52

Length of c a s l n g s t r l n g feet.. ...........................

Calcula ted weight In a l r of s tr ing, pounds ............

Calcula ted welght In a l r of c a s i n g below t e s t point, pounds ................

A P I rnlnlrnurn jolnt pull-out s t rength , pounds* ...........

Calcula ted t e n s l o n d e s l g n factor ..........................

Depth of t e s t polnt from sur face , fee t .................

Hesul t ........................... Average mud welght , pounds

..................... per gal lon .............. Bouyancy fac tor

Bouyant welght of c a s i n g s t r lng , pounds ...............

Bouyant w e ~ g h t of c a s l n g below t e s t polnt, pounds..

Increased load d u r ~ n g ........ cementing, pounds G

T o t a l t e n s i l e load a t t e s t ................ p o ~ n t , pounds

Calcu la ted d e s l g n factor a t t e s t point d u e t o bouyant load recipr ocatlon.. .........

T e s t No. 2 G. W.

McKenzle No. 5 8-7-52

10,133

161,604

136,184

21 1,000

1.55

1,551 t

9.35 0.857

138,495

116,709

26,000

142,709

1.480

T e s t No. 3 W. H.

Fender "A" No. 2

8-17-52

10,695

172,968

140,668

211,000

1 .50

1,902 t

9.30 0.858

148,406

120,693

30,000

150,693

1.401

T e s t No. 4 C. 1.

Jarrel l - N o . 2 8-28-52

10,380

167,528

. -

140,778

211,000

1 .50

1,573 t

9.15 0.859

143,906

120.928

3 1,000

151,928

1.390

T e s t No. 5 E.

Hutton "B" No. 3

10-4-52

11,020

178,741

145,741

211,000

1.45

1,946 t

9.40 0.856

153,002

124,754

29,000

153,754

1.373

T e s t No. 6 T e s t No. 7 H. H. B.

Kelly G ~ b b o n s "D" No. 3 "B" No. 1 10-12-52 10-12-52

* From API Be1 5C2 for 5'/2-,n., 15.5 Ib per ft, 5-55. STC casmg. t Successful. GIncreased load due to recrprocat~on, w e l g h t of cement, etc.

would be generally lower than those calculated when these factors are neglected. Collapse Tests

Testlng collapse deslgn factors was much more complicated than t e s t ~ n g tension d e s ~ g n factors. A sketch of the method used, Fig. 3, i s included and the procedure used is briefly described a s follows. In each test the portion of the casing string above the tes t section was designed in our usual manner. The three lengths of casing to be collapse-tested were run on bottom of thls string and collapse tes ts were obtained by conducting drill-stem tes ts ~ n s ~ d e the casing. In every case the collapse-tested section of casing was below all commercially productive zones, and the use of a flapper-type float shoe without a float collar permitted the casing string to be drifted to bottom after each successive collapse test . Every strlng was equipped with centralizers and reciprocating scratchers in the manner previous-

T e s t No. 8 G.

b l ~ k l e s No. 4

12-3-52

10,600

170,323

150,693

211,000

1.40

1,154 t

9.10 0.861

146,648

129,746

35,000

164,746

1.281

l y described. After the casing was run In the well ~t was landed on s l ips and dr~f ted to bottom with a drilling bit run on tublng. The caslng was a lso dr~f t - ed w ~ t h a b ~ t after each successive collapse test .

Collapse tes ts were obtained on both uncemented and cemented caslng by conduct~ng drlll-stem tes t s Inside the test section. The test-tool packer was s e t In the last length of casing of the normally designed s t r ~ n g , and the diflerentlal collapse pressure desired was controlled by varylng the aniount of water cushion placed in the tublng. Collapse d e s ~ g n factors for the test s e c t ~ o n , calculated from the differential between external mud-colun~n pressure and internal water-cushlon pressure, were progressively lowered In increments of 0.05 until the casing collapsed or until a tes t was obtained with zero internal pressure. All caslng used In the collapse-tested section was presswe tested to 80 percent of mini- muni yield pressure (4,400 ps i for 5'4-in., 15.5-lb,

F I E L D TESTING O F CASING-STRING DESIGN FACTORS 27

I?N.B= 1 7 w J . 5 L L E E r -

#5 ~ ' J x ~

BASE SALT STRING

I5 5*J-55 STC

6.700

STRING

15 5 5 - 1 3 STC

DESIGN FACTMI-TENSION = I 60 DESIGN FACTOR-TENSION ' I 4 0 DESIGN FACTOR-BURST = I 33 DESIGN FACTOR-BURST - 1 33

DESIGN FACTOR-COLLAPSE =I DO~UNCEMENTED) DE5lGN FACTOR-COLLAPSE I OD ~UNCEUENTEO) DESlCH FACTOR COLLAPSE *B~(CEUENTCD) DCSlCN FACTOR-COLLAP.51-0.3 (CEMENTED)

NOTE DESIGN FACTOR FOR BURST BASED ON 4.000 PSI WELL HEAD PRESSURE

Fig. 2-11,000-ft Well-5lkin. Casing, 9A Ib per gal Mud

3-55 casing). Also, the weight, d r~f t , minimum wall thickness, and eccentricity of each length was checked. Thls lnformat~on, together with data from

-

bloody Engineering Co. lnspect~on reports, is tabulated i n Tables 2 and 3.

In all, five collapse tests have been conducted and the results have been tabulated I n Tables 4, 5, 6, 7, and 8. A summary of these results 1s a s follows:

Design Factor* at which

Collapse Test Design Factor Collapse No. Tested Occurred

1 (uncemented) 0.95,0.90,0.85,0.80 0.80 2 (uncemented 0.95 0.95 3 (uncemented) 0.95, 0.90, 0.85 0.85 4 (uncernented) 0.95, 0.90 t

(cemented) 0.85, 0.80, 0.75 t 5 (uncemented) 0.90 t

(cemented) 0.75,0.70,0.60,0.50 t *Note that the effect of tenslon on collapse 1s not considered

because bottom joznts have no tenslle load. t Caslng dld not collapse.

All of the foregoing collapse design factors were calculated from AFI nllnlmunl collapse strengths. For convenience, the collapsing force for both uncemented and cemented caslng was assumed to be that imposed by the dlllerential between a full external mud-column and water-cush~on pressure. The valid~ty of thls assunlption IS discussed later here- In. Although collapse failure was experienced on three of the five tests, the well casing was successfully cemented through perforat~ons on the occasion of each fallure. However, i n all cases the caslng

2 i h~ . ~ ~ , * E u E , J-55 TUBING

Fig. 3-Diagram of Casing Collapse Test

28 J. E . SAYE AND T. W. G . RICHARDSON

Table 2 Fleld-inspection Results for Collapse Test Casing

Length Threads Stenclled Actual Calculated

Collapse Test Jolnt Drlft Off, Weight, Weight, Feight, No. No. Test Feet Pounds Pounds Lb per Ft.

f

E. Hutton "13" ~2 O.K. 31.45 500 502.5 15.8 5'4-ln. OD, 15.5-lb,

I

1 O.K. 31.71 502 511.0 16.1

E. C. Kann # 4 5'4-ln. OD, 15.5-lb, { 2 O.E. 31.56 5 00 501.5 15.9

1 O.K. 31.91 508 511.0 16.0

'8. hl. Rumberger # 3 { 2 0 . R . 31.81 5 04 504.5 15.8 5'4-in. OD, 15.5-lb,

O.K. 32.91 520 525.5 16.0

1 O.K. 30.58 484 494.0 16.2

L. E. Jones # 1 5'4-ln. OD, 15.5-lb, ( 2 0 . K . 30.65 486 506.0 16.5

Blackrnon-Woody # l 2 0 . ~ . 32,36 453 456.5 14.1 5'4-in. OD, 14-lb,

11-40 Casing I C

* Wall-thickness read~ngs were taken 120 deg apart on the pm end. t OD readlngs were taken approximately 5 ft from pln end, mld-sectlon, and 5 ft from collar.

Cal~pered Wall

Thickness, Inches* 0.28 0.27 0.27 0.27 0.27 0.27 0.27 0.26 0.29 0.28 0.28 0.28 0.27 0.27 0.26 0.26 0.27 0.25

Callpered OD, Inches t -

Max. klln. 5.50 5.49 5.50 5.50 5.49 5.49 5.49 5.48 5.49 5.19 5.50 5.49 5.49 5.49 5.49 5.49 5.49 5.49 5.50 5.47 5.50 5.47 5.49 5.47

was frozen and could not be reciprocated. This caused difliculty In cenientlng and several expen- sive squeeze-cementing jobs were needed to obtaln a good cement sheath behind the plpe.

It will be noted froni Tables 4 and 5 that, in the first two collapse tes ts , calculated design factors were slightly diiterent from those indicated by the two pressure recorders in the test tool. In reviewing the procedure used and results obtained on these initla1 tes ts , it was concluded that erroneous mud- column and water-cushion pressures were recorded. The error in recorded mud-colunln pressure was a result of piston ellect In running the packer and packer "squeeze." This procedural inaccuracy was reduced In subsequent tes ts by permitting pressure above and below the packer to equalize before the packer was s e t and by waitlng 30 nlin before open- Ing the tes t tool. Typical bottom-hole pressure charts are ~l lus t ra ted in Fig. 4 and 5. Another source of error, especially noted on recorded water-cushion

Table 3 Specificat~ons for Collapse-tested Cas~ng

Collapse Collapse Tests Test

No. 1-4 No. 5 Outslde diameter, ~nches, nominal.. 5.5 Weight per foot, pounds, nomlnal .... 15.5 Grade ...................................... J-55 Threads ................................. 8-rd, short

.................................... Range 2 blanufacturer ........................... A

Seamless Conversron

Mill test pressure, psi ................ 3,000 Field test pressure, psi .............. 4,400 Dlameter of 6-in. long drlft mandrel, Inches ....................... 4.825

Average physical properties:* Number of tenslle tests ............. 15 Average tensile strength, psl ...... 102,702 Average yleld strength, psi ........ 63,115 Minlmum yield strength, psi ........ 60,400 Elongation, percent In 2 in. ........ 26.27

5.5 14.0 H-40

8-rd, short 2 B

Seamless 3,000 2,800

4.887

3 85,550 51,080 49,140 33.67

*From ~ n s p e c t ~ o n report b y Moody Englneerlng Company.

Desired deslgn fac tor1 . . ............... 0.95 0.90 0.85 Packer setting depth, fee t . . ............ 10,538 10,538 10,532 Tabulated niininiuni collapse strength, psi. 3,860 3,860 3,860 Calculated collapse strength, p s i 2 . ...... 4,780 4,780 4,780 Mud weight, lb per gal ................. 9.35 9.40 9.35 Calculated hydrostatic mud pressure, psi . 5,110 5,143 5,108 Water cushion, f e e t . . .................. 2,471 1,940 1,270 Calculated water-cush~on pressure, p s i . .. 1,070 840 550 Calculated pressure difterentlal, psi ..... 4,040 4,303 4,558 Calculated design factor' .............. 0.955 0.897 0.847 Type pressure recorder ................ Bomb 1 Bomb 2 Bomb 1 Bomb 2 Bomb 1 Romb2 Recorded hydrostatic mud pressure, psi . . 5,149 5,200 5,081 5,150 .- 5,150 4,925 Recorded water-cushion pressure, ps i .... 1,134 1,140 932 990 625 700 Recorded pressure diHerentia1, ps i ...... 4,015 4,060 4,149 4,160 4,525 4,225 Indicated design factor ............... 0.961 0.951 0.930 0.928 0.853 0.914 Result ............................... T e s t Tes t Tes t

Successful Successful Successful ' Baaed on tabulated m n l m u m collapse values.

Based on mlnlmum yleld atrength f rom Moody lnspectlon report. Calculated value based on water cush~on, preesure recordere felled to record.

pressures, was rat~onalized by an investigation of instrun~entation limitations. It was deternilned that the magnitude of errors noted corresponded closely to the lnherent l imitat~ons of the instrunlents used, these lini~tations being due priniarlly to spring h y s - teresis and calibration. The use of the foregoing procedural changes, carefully selected'lnstrunrents, and carefully read data on later tes ts resulted in

niuch closer agreement between calculated and nleas- ured values. The magnitude of di~ierence was then considered wlthin the range of experimental accu- racy.

In the first collapse tes t the casing collapsed a t a designfactor of 0.79 after successfully' wlthstand- Ing tes ts a t design factors of 0.95, 0.90, and 0.85. On tes t no. 2 the pipe collapsed a t a design factor

Table 4 Cas~ng Collapse Test No. 1

E. Hutton "B" No. 2-July 13-16, 1952 Elk City F ~ e l d , Beckham County, Oklahoma

30 J. E. SAYE AND T. W. G. RICHARDSON

Table 5 Caslng Collapse Test No. 2

E. C. Wann No. 4-July 26, 1952 Elk C ~ t y F ~ e l d , Beckham County, Oklahoma

Desired des~gn factor' ........................ 0.95 Packer setting depth, feet .................... 10,382 Tabulated minlmum collapse strength, PSI 3,860

......... Calculated collapse strength, psi2 4,780 Mud Weight, Ib per gal ....................... 9.20 Calculated hydrostatic mud pressure, P S I . 4,963 Water cushlon, feet ............................. 2,082 Calculated water-cush~on pressure, PSI ... 904

....... Calculated pressure different~al, ps i 4,059 Calculated des~gn factor' .................. 0.951 Type pressure recorder ....................... Bomb 1 Bomb 2 Recorded hydrostat~c mud pressure, PSI ... 5,125 5,203 Recorded water-cush~on pressure, p s i ..... 904) 9 w 3 Recorded pressure d~fferential, PSI. ....... 4,221 4,299 Indicated deslgn factor ....................... 0.914 0.898 Result .............................................. Casing

Collapsed ' Based on tabulated mrnrrnum collapse values. 'Based on rnrnlmum yreld strength f rom Moody lnspectron report.

Calculated value based on w a t e r cushron, pressure recorders felled to record.

of 0.95. C a s q used In test no. 3 successfully withstood design factors of 0.95 and 0.90, but collapsed at 0.85.

After studylng the results of the first three tests, the fourth test was successfully conducted on unce- n~ented casing at design factors of 0.95 and 0.90.

Date. 9-22-52 Mud We~ght. 9.3 Ib per gal Total Depth: 10,788 ft Water Cush~on: -2,059 ft Packer Set At 10,665 ft Top Choke Stze: tn

Recovery. Water Cush~on and 100 ft of mud Fig. 4-Sample Drill-stem Test Chart for Uncollapsed

Casing Casing Collapse Test for 0.90 Design Factor

Shell Oi l Co: W. M. Rumberger No. 3 Elk City Field

Table 6 Cash-ng Collapse Test No. 3

W. M. Rumberger No. 3-September 21-23, 1952 Elk City F ~ e l d , Beckham County, Oklahoma

Des~red des~gn factor1 .................... 0.950 0.900 Packer setting depth, feet. . ............... 10,656 10,655 Tabulated nllnlnluni collapse strength, psl ... 3,860 3,860 Calculated collapse strength, ps12 ......... 4,780 4,780 hlud we~ght, Ib per gal .................... 9.40 9.35 Calculated hydrostatic n~ud pressure, psi . . . . 5,200 5,180 Water cushion, feet . . ..................... 2,639 2,059 Calculated water-cushlon pressure, psi ..... 1,145 894 Calculated pressure ddlerential, psi . . . . . . . . 4,055 4,286 Calculated design factor ................. 0.952 0.901 Type pressure recorder ................... Bomb 1 Bomb 2 Bomb 1 Bomb 2

..... Recorded hydrostatic nlud pressure, psi 5,217 5,250 5,243 5,300 Recorded water-cushion pressure, psi. ...... 1,099 1,250 M1 1,075 Recorded pressure dillerentlal, psi ......... 4,118 4,000 4,272 4,225 Indicated d e s p factor' .................. 0.937 0.965 0.904 0.914 Result.. ................................ Test Successful Test Successful

0.850 10,654 3,860 4,780

9.35 5,180 1,479

64 1 4,539 0.850

Bomb 1 Bomb 2 5,186 5,350

641 ' 641 4,545 4,709 0.849 0.820 Casing Collapsed

'Based on tabulated m n r m u m collapse values. 'Based on mlnrmum yreld strength f rom Moody rnspection report. ' Calculated value based on water cushlon. pressure recordrng unreadable.

Table 7 Cas~ng Collapse Test t4o. 4

L . E. Jones No. 1-November 19, 1952 Elk C ~ t y Field, Beckham County, Oklahoma

Uncemented Cemented' PI

I , / \ Desired design factor2 ................. 0.95 0.90 0.85 0.80 Lowest

Possible Packer setting depth, feet ............. 10,510 10,508 10,506 10,504 10,502 Tabulated minimum collapse

strength, psi ............................. 3,860 3,860 3,860 3,860 3,860 Calculated collapse strength, ... 4,780 4,780 4,780 4,780 4,780 hlud welght, I b per gal .................. 9.4 9.4 9.4 9.4 9.4 Calculated hydrostatic mud

pressure, psi ............................ 5,130 5,130 5,130 5,130 5,130 Water cushion, feet ...................... 2,472 1,924 1,346 706 0 Calculated water-cushion pressure,

psi .......................................... 1,070 83 4 5 83 3 06 0 Calculated pressure differential, psi 4,060 4,296 4,547 4,824 5,130 Calculated design factor2 .............. 0.95 0.90 0.85 0.80 0.75 Calculated design factor after

....... cementation4 ............................. ....... 0.951 0.891 0.832 Type pressure recorder ................. Bomb 1 Bomb 2 Bomb l Bomb 2 Bomb l Bomb 2 Bomb l Bomb 2 Bomb l Bomb 2 Recorded hydrostatic mud pressure,

PSI ......................................... 5,166 5,200 5,123 5,060 5,130' 5,130' 5,130' 5,130' 5,130' 5,130' Recorded water-cush~on pressure,

psi ........................................ 1,129 1,100 883 800 626 590 331 340 34 0 Recorded pressure differential, psi .. 4,037 4,100 4,240 4,260 4,504 4,540 4,799 4,790 5,096 5,130 Indicated design factor2 ................ 0.95 0.94 0.91 0.91 0.86 0.85 0.81 0.81 0.76 0.75 Result ....................................... Test Test Test Test Test

Successful Successful Successful Successful Successful ' The pressure on the outslde of the caslng after cementation was assumed to be equal

to the pressure prior to cementing for purpose of calculating deslgn factors. Based on tabulated mrnlmum values.

'Based on m~nnnum yleld strength from Moody lnspectlon report. D e s ~ g n factor of collapse calculated uslng pressure on outslde of caslng based on formation pressure equal to the we~ght of salt-water column welghlng 0.442 p s ~ per foot of depth.

A s the t e s t s e c t i o n w a s not co l lapsed , the c a s l n g w a s cemented In the normal manner and three t e s t s were conducted on the cemei~ ted c a s l n g in a n a t t e m p t t o eva lua te the effect of cJement behind the pipe on c a s l n g co l lapse . If calculated on the same assump- t ~ o n as for uncemented cas lng , d e s l g n fac tors on t h e s e t e s t s were 0.85, 0.80, and 0.75 and the pipe w a s not col lapsed. No water c u s h ~ o n w a s used in the l a s t of th i s s e r l e s of t e s t s , which, based on calculated mud-column w e ~ g h t , would correspond to a des lgn factor of 0.75. However, ~f a f te r cementing, the external pressure on the pipe i s Instead equa l t o formation pressure, the design factor would be 0.83 if a formation pressure gradient of 0.442 p s i per foot is assumed.

T o check the val idl ty of r e s u l t s o b t a ~ n e d in t e s t no. 4, another co l lapse t e s t w a s conducted. In th l s t e s t three lengths of 5%-in. OD, 14-lb, H-40, STC c a s i n g were used In the t e s t sec t ion , whereas pre- v ious t e s t s had used 5%-in. OD, 15.5-lb, J-55, STC cas lng . T h l s H-40 c a s i n g success fu l ly withstood

co l lapse pressures corresponding t o a co l lapse de- s i g n factor of 0.90 when uncemented, a n d design fac tors of 0.75, 0.70, 0.60, and 0.50 after the cas - lng w a s cemented. However, for convenience, a l l d e s i g n fac tors were aga ln ca lcu la ted on the b a s i s of ful l mud-column pressure outs ide c a s i n g and water-cushion pressure Inside, even though pipe w a s cemented for the l a s t four t e s t s . There IS some doubt that cemented c a s l n g w a s ac tua l ly co l lapse tes ted . Our reason for quest ioning t h e s e r e s u l t s 1s t h a t the zone opposl te the cemented col lapse-test s e c t i o n w a s of s u c h low permeab~l i ty that fornla- t ion pressure might not have been suflicient t o ob- t a in the design fac tors mentioned.

. In the wel l used for t e s t no. 5, th i s co l lapse- tes t s e c t i o n w a s located between 10,755 and 10,852 f t In a n interval h a v ~ n g no permeability of commercial slgnificarlce. E v e n s o , the zone of highest porosity and permeability i n t h ~ s s e c t i o n of the hole had been open-hole drill-stem tested In the interval froni 10,812 t o 10,827 ft. T h l s open-hole t e s t recorded

J. E . SAYE AND T. W. G. RICHARDSON

Table 8 Casing Collapse Test No. 5

Blackmon-Woody No. 1-June 1-7, 1953 Elk C ~ t y Field, Beckham County, Oklahoma

Uncemented Cemented' h

.... Deslred collapse design factorZ w r 0.75 0.70 0.60 0.50 ' Packer settlng depth, feet ........... 10,728 10,726 10,723 10,721 10,719 Tabulated minlmum collapse

strength, p s ~ 3 ........................ 2,440 2,440 2,440 2,440 2,440 Calculated collapse strength, ps~4. . 3,195 3,195 3,195 3,195 3,195

................. blud weight, lb per gal 9.3 9.3 9.3 9.3 9.3 Calculated hydrostatic mud

...................... mud pressure, psi 5,180 5,180 5,180 5 , 1 8 0 5,180 Water cushion, feet ................... 5,704 4,440 3,906 2,557 6 74 Calculated water-cushion pressure,

PSI ....................................... 2,470 1,925 1,690 1,108 2 92 Calculated pressure differential, psi 2,710 3,255 3,490 4,072 4,888

.............. Calculated design factor 0.90 0.75 0.70 0.60 0.50 Calculated d e s ~ g n factor after

cementin g5... ........................... ....... 0.867 0.811 0.672 0.549 Type pressure recorder ................ Bomb 1 Bomb 2 Bomb 1 Bomb 2 Bomb 1 Bomb 2 Bomb 1 Bomb 2 Bomb 1 Bomb2 Recorded hydrostatic mud

........................... pressure,psi 5,163 5,175 . 5,180 5,180 5,180 5,180 5,180 5,180 5,180 5,180 Recorded water-cushlon

........................... pressure, PSI 2,451 2,500 1,914 2,000 1,706 1,750 1,100 1,200 335 300 Recorded d~fferential pressure, psi.. 2,712 2,675 3,266 3,180 3,474 3,430 4,080 3,980 4,845 4,880

............... Indicated design factorz 0.90 0.91 0.75 0.77 0.70 0.71 0.60 0.61 0.50 0.50 Result .................................. Test Test Test Test Test

Successful Successful Successful Successful Successful ' For purpose of calculating d e s ~ g n factors pressure outs~de caslng after cementlng

assumed to be equal to outslde pressure pr~or to cementlng. Based on tabulated API mlnlmum strength. Mln~mum collapse resistance for 5'4-ln., 14-lb, H-40 caslng from API Bul 5C2. Based on mlnlmum y ~ e l d strength from Moody lnspect~on report. Collapse d e s ~ g n factor calculated on mlnlmum strength value assumlng formatlon pressure outs~de caslng to be that of salt-water column we~ghlng 0.442 p s ~ per foot.

a bottom-hole f o r n ~ a t ~ o n pressure of only 1,010 ps i , both open and shut-in, indicating no entrance of res- ervolr fluid. Therefore, ~t w a s our opinion that cemented c a s i n g in the t e s t sec t ion of t h i s wel l nright not have been sub jec ted t o a co l lapse pressure greater than that corresponding t o a des ign factor of approximately 3.0. Our reason for c o n s ~ d e r i n g t h l s factor l i e s in the belief that cement ef lect ively bridges the annulus between cas ing and open hole, if the cement job is good, r e l ~ e v i n ~ cemented c a s - ing of external pressure caused by the mud co lun~n, thereby only subjecting cemented cas ing t o co l lapse fo rces due t o fo rmat~on pressures . Because it i s doubtful that a n Impermeable formation could exert a col lapsing force, s u c h forces c a n be appl ied on c a s i n g and cement only by porous, permeable forma- t ions having pressures sufliciently greater than internal cas lng pressure.

In view of the foregoing d iscuss ion , the d a t a on open-hole drill-stem t e s t s conducted in col lapse- t e s t wel l s have been tabulated in T a b l e 9. It i s t o

be noted that not one of these w e l l s recorded a for- n ~ a t ~ o n pressure on d r ~ l l - s t e m t e s t which w a s sufli- c i e n t t o co l lapse cas ing . However, it must be remembered that these open-hole drill-stem t e s t s were of short duration and it i s p o s s ~ b l e that true formation pressure might not have been recorded because of the t ~ g h t n e s s of the formation. If a l lowed t o s t a - bi l ize , s u c h zones might, In time, build up sufficient pressure to sub jec t the cas lng t o co l lapse .

B e c a u s e of our thoughts concerning the co l lapse fo rces present af ter c a s i n g w a s cemented i n t e s t no. 5, we have planned another t e s t for a well i n w h ~ c h formation o p p o s ~ t e the col lapse-test s e c t i o n wil l be drill-stem tes ted t o prove the ex i s tence of fo rmat~on pressure greater than tabulated col lapse- s t rength pressure and t o show tha t the zone h a s sufficient permeabi l~ ty to transmit t h i s pressure quick- ly . We have proposed tha t co l lapse t e s t no. 6 wil l b e conducted exact ly in the s a m e manner as co l lapse t e s t no. 5 except that c a s i n g wil l be cemented before any co l lapse t e s t s a r e made. Cemented c a s i n g

will then be collapse-tested to design factors of 0.75 and lower, ~f possible. The casing to be used for thls test will be three lengths of 5'4-in. OD, 14 lb per ft, H-40, STC casing w h ~ c h has been inspect- ed by bloody Engineer~ng Con~pany. Selection of the well for tliis test will depend upon having a porous, permeable zone opposite the collapse-test section whlch has a formation pressure of 3,250 psi or greater.

En~ergency Casing-stnng Design During June 1952, a t the same tinre that planned

tes ts of tension and collapse design factors were b e ~ n g conducted, the shortage of 5'4-in., N-80 caslng became s o acute that ~t was necessary to devi- ate froni our standard casing-string design in order to obtaln sufliclent caslng strings to niaintaln the drllllng program In the Elk City F'leld. A s an expe- d ~ e n t to reduce requiren~ents for N-80 casing, an emergency cas~ng-string design was adopted on a tenlporary basls. By reducing the collapse design factor from 1.00 to 0.85 In the cemented portion of the caslng strlng, it was posslLle to subs t~ tu te In each well approximately 1,500 ft of 5'4-111. OD, 17 1L per ft, 5-55 caslng for the N-80 caslng normally requwed. Uslng this procedure with our usual cas- ~ng-string design metllod allowed us to design 10,800- ft strings of 5'4-ln. caslng w~thout any N-80 plpe.

Date: 9-23-52 Mud Weight: 9.35 Ib per gal Total Depth: 10,788 ft Water Cushton: -1,479 ft Packer Set At: 10,654 ft Top Choke S ~ z e 3/,, In

Recovery. Water Cushron and 100 ft of mud Fig. 5-Sample Drill-stem Test Chart for Collapsed

Casing Casing Collapse Test No. 3 for 0.85 Design Factor Shell Oil Co. W. M. Rumberger No. 3 Elk City F ie ld

L. 0 mP- E

cow- 2 % , a E P-I-- C 0 s 2 dlddp o o m : - ~ $ $ E B o w

2,,1n u m q u z a e l $ ; * m - E L , - = g m - . % E g 2 9 +-: 5

- " 2 2-+ - i S ," + In 0) L TI 0 = 2 z + m il

2 5 z 2-5 *-

" a

0 1 4 - 0: o a c u m o % I n 0 2 G O O

0 - - &P- - -S 'o *uo ;mmm-o-;mm-

corn- m k t - m--gL= $ I n 0 c O * z m M U E

m m

E I B 2 -

DEPTH IN FEET 9500 10000 10500 11 000 11500

Mud We~ght: 9.4 Ib per gal Des~gn Factor, Tens~on: 1.40 D e s ~ g n Factor, Collapse: As ~ n d ~ c a t e d Des lgn Factor, Burst: 1.33 at 4,000 P S I

we1 lhead pressure F ig . 6-Special 5'4-in. Casing-string Design Chart

Elk City Field, Beckham County, Oklahoma

Strings of this design have been used for over 18 -

months and no fa~ lu res or other dilticulties have been encountered. T h ~ s method has been adopted a s standard for the Elk C ~ t y F ~ e l d .

Although the emergency design was not originally a part of the program for field testing casing-string deslgn factors, the experience obtained through ~ t s use cannot be ignored. In using thls design, we have been careful to obtain ,good cement fill-up behind the pipe, at least up to the point where the d e s ~ g n

- -

factor In collapse is 1.00. \be have not, however, found ~t necessary to squeeze-cement any of these strings.

SUMMARY In summary, there was a threefold reason for ini-

tlatlng the t e s t s of casing-str~ng design factors: 1, shortages of tubular goods; 2, designs of casing

-

strings for greater depths; and 3, econonlles which might be reallzed. Even though the shortage of tubular goods is no longer acutk, the results of this In- ves t~gat ion are considered lnlportant because design of deep-well casing is gea t ly facilitated. .4vailable

grades of casing can be used to better advantage when design factors are more realistic. The econo- mles ach~eved by utllizlng the reduced design factors of t h ~ s paper, although not spectacular, do con- tribute to reducing well costs.

A s a result of successful experience presented herein, Shell Oil Company has adopted the following design factors a s standard for the Elk City Field: 1.40 in tension; 1.00 in collapse for uncemented casing; and 0.85 in c,ollapse for cemented caslng. Be- cause no test work was done with burst, the burst design factor of 1.33 rernalns unchanged. Addit~onal credence is glven the test results because 40 strlngs

-

of the foregoing design (Fig. 6) have been run in Elk C ~ t y Field wells without any d~lliculty whatever.

The authors believe that the results of tension t e s t s are v a l ~ d and allow design-factor reduction from 1.60 to 1.40 when the Shell design method i s

-

used. Collapse t e s t s of cen~ented casing, although needing the clarlficafion which wlll come from the addi t~onal test that has been planned, are s u l l ~ c ~ e n t - ly encouraging to allow the use of a 0.85 collapse design factor in the cen~ented portlon of the caslng s t r ~ n g . Because of the spread of design factors a t whlch u n ~ e m e n t e d ~ c a s l n ~ encountered collapse failure, ~t IS believed that nlaximum ut~lization of caslng IS being obtained w ~ t h a design factor of 1.00 in uncemented s e c t ~ o n s of the strlng. The results of the tes ts reported do not warrant reduction of thls deslgn factor.

ACKNOWLEDGEMENT The program for field testing casing-strlng design

factors grew out of Shell Oil Cor~~pany's long expe- rlence wlth cas~ng-string d e s ~ g n , and consequently the authors are deeply indebted to this company and the Inany engineers who contr~buted to the develop- ment of design methods. I'rorn~nent anlong these engineers were RI S. Cralie and Laurence O'Donnell. Also, the wr~ te r s w ~ s h to express their appreclatlon to Shell 011 Company for accepting the expense and risk lnvolved In conducting the field t e s t s and for grant~ng permission to publlsh the data presented he re~n .

In addition we gratefully acknowledge the ass is t - ance of the following Shell engineers In preparing this paper: T. I!. Dwyer for his advice and suggestions. J . D. Goodrlch, who first proposed the field-test

progranl,and, with the ass is tance of li. E. Fink, did the basic planning of it.

I;. L. Hankin, who prepared the first progress re-

port on the program. C . K. l ie i ter and F. L. Rloore, who observed the

t e s t s in the field and compiled field da ta . l ' h e authors a l s o received a id and advice from

H. G. l 'exter , Spang-Chalfant D ~ v i s ~ o n of T h e Na- t ional Supply Company, and Arthur hladdox, l'he Carter 011 Company.

BIBLIOGRAPHY Heports of the Committee on Standardrzation of 011-Coun- try Tubular Goods, Proc. Am. Pet Inst. (Prod. Eul 201- 237) (1928 through 1951).

Drrllrng and Productron Practrce, American Petroleum In- stitute, 1935-1952.

API Bul 5C2. Performance Propertres of Caslng and Tub- rng, Dallas, January 1952, Flfth Edition.

Engrneerrng Data, Spang-Chalfant Dlv. of The National Supply Co., July 1952 (Rev )

Booklet No 60, Youngstown Sheet and Tube Co , 6th Edi- tion, 1950.

Payne, John M. A Study Group lnvestigatlon of Equipment and Techniques for 20,000-ft Drilling, Drrllrng and Pro- ductron Practrce, 123 (1949).

O'Donnell, Laurence, and Crake, W. S: Mechanical Causes of Casing Fallure and Practices for Theu Control, 011 Gas J , Dec. 16 (1943).

Hllls, J . 0. A Review of Casing-strlng Deslgn Principles and Practice,Drrllrng and Productron Practrce, 91 (1951).

Cooley, H. M . Caslng-selection Charts, Petroleum Engr- neer, 16 [4] 241 (1945).

Cooley, H. M. Preparation and Use of Casing-selection Charts, Or1 Weekly, 113, 10 (1944).

Curran, B. E. Combination Caslng Strlngs, 011 Gas ] , 80, Oct. 14 (1945).

Stewart, Reid T Collapslng Pressure of Lapwelded Steel Tubes, Am. Soc Mech Engrs., May, 1906.

Jasper, T. blclean, and Sullivan, J. W. W: The Collapslng Strength of Thln Tubes, Trans. Am Soc. Mech Engrs. (Applred Mechanrcs) 53, 17b (1931).

Rescott, B. B, Dunlop, C. A, and Kemler, E. N Se t t~ng Depths for Casing, Drrllrng and Productron Practrce, 125 (1940).

Holmquist, J . L. and Nadai, A: A Theoretical and Ex- perimental Approach to the Problem of Collapse of Deep- well Casing, Drlll~ng andProductron Practrce, 392 (1939)

Edwards, S. H. and Miller, C. P Discussion of the Effect of Combined Longitudinal Loading and External Pres- sure on the Strength of 011-well Caslng, Drrlllng and Productron Practrce, 483 (1939)

Kettenburg, R . J. and Schmleder, Fremont H 011-well Casing Failures, Drrllrng and Productron Practrce, 185 (1945).

Texter, H . G Casing Straln after Cementing, 011 Gas I . , Aprll 8 (1948).

Lubinsk1,Arthur: A stud; of the Buckllng of Rotary Drlll- lng Strlngs, Drrllrng and Productron Practrce, 178 (1950).

Kl~nkenberg, A. The Neutral Zones In Drlll Plpe and Cas- lng and Their Significance in Relat~on to Buckllng and Collapse, Drrllrng and Productron Practrce, 64 (1951).

Cl~nedlnst, Vi. 0. Collapse Safety Factors for Tapered Casing Strlngs, Drrll~ng and Productron Practrce, 181 (1945).

DISCUSSION H. R'. Ladd (Stanolind 011 and G a s Co., T u l s a )

(wr~tten): This paper excel lent ly p resen ts the r e s u l t s

of a well-thought-out and executed program of testlng the joint s t rength and co l lapse properties of cas - ing under controlled c o n d ~ t i o n s of s t r e s s . I'he re- s u l t s a re of much interest and inlportance t o a l l pe- troleunl producers,and l t is certain that th i s presen- ta t ion by Shell through the API wil l be greatly ap- preciated.

I quite agree w ~ t h the opinion expressed tha t the des ign safety factors for cas ing tension and col- l apse now generally appl ied a g a i n s t API mininlunl perfornlance properties c a n be lowered without ex-

- .

c e s s i v e r l sb of f d ~ l u r e . In the la te 19'30's, my company experinlented with

the lowering of des ign fdctors , and by progresswe tr ia l arrived a t 2.00 for joint s t rength in tension and 1.50 for co l lapse res i s tance based on the old aver- a g e perforn~ance properties of cas ing . T h e s e were converted t o 1.60 and 1.125 when the b a s i s for e s - tabl lshing API m ~ n i n ~ u n l properties w a s adopted in 1940. T h e s e a r e not far from the factors Shel l w a s using, and we a l s o had no fai lures attributable t o design.

l ' h e AF'I performance properties adequately re- present the mlnimuni v a l u e s of c a s i n g and c a n be confidently used as a b a s i s for design. T h l s h a s been proved in the pas t and i s further subs tan t ia ted by th i s paper. Re s topped a t a co l lapse factor of 1.125 for nornlal des ign , although we fel t that , ~f the minimum AFI co l lapse va lues were accurate , a factor down t o 1.000 could be sa fe ly used; and we did go toward that limit when we were forced t o ap-

-

ply it. T h ~ s paper p resen ts s t rong evidence tha t the use of a co l lapse factor of 1.00 is pract ical for un- supported cas ing under normal conditions. \\hen un- usual c o n d i t ~ o n s a re expected, s u c h a s abnorn~a l external pressure, heavy s w a b b ~ n g , e tc . , the operator nlight run a d e f i n ~ t e r l sk of co l lapse that would ne- c e s s i t a t e a ra i s ing of t h i s d e s i g n factor t o a degree that would afford more protection.

T h e t e s t s made and proposed t o de te rn~ine the ef- fec t of cement In res i s t ing co l lapse pressures bring a n important developnient into casing-string design practices. Heretofore, many have considered that t h i s etiect should not be a part of design, presunia- bly because of fear that cementing could not be suf- f i c~ent ly controlled. biodern cementing pract ices and a c c e s s o r i e s have gone a long way toward insuring good cement jobs, and the r i s k of failure i s small. T h e t e s t s i n d ~ c a t e a posi t ive b a s i s for evaluat ing the oflsetting ellect of cement on col lapse pressures and show tha t the final determinations wil l provlde

36 J. E . SAYE AND T. W . G. RICHARDSON I

a practical basis for il~aklng an adjustn~ent of the collapse factor

Although the burst design factor was not field- tested, it IS n~entioned in the paragraph on casing- string design n~ethods although it is not clearly In- dicated a s to how it IS applied. It is iniportant to distinguish whether the factor for Internal pressure is applied on the nrlnlttlum yield strength or on the niininlunl t e n s ~ l e strength of the caslng. Dependtng upon the g a d e of casing, a factor of 1.00 on the yield strength is a [actor of 1.25 to 1.50 on t l ~ e ten- s i le strength. l 'he use of the nlininiunl internal yield pressure in deslgn seems preferable, and we have cons~dered a factor of 1.00 on thls bas is to be adequate.

It IS tremendously Interesting to observe the resu l ts of the tension tes ts whereby the design factor is lowered to 1.40. \\e have used 1.60 for a long

-

tlnie and have had sonie qualnrs a s to whether it was too risky, but have had no pull-outs by reason of uslng this factor. l ' h i s paper, a t the least , gives us assurance that our present factor is adequate, and furthern~ore prov~des a defin~te basls for cons~dera- tion of reducing it. I do want to call attention to the fact that, in dropplng the tension factor to 1.40, the nlinlnlunl yield strength of the area under the last perfect thread or gage point is being approaclled in sonbe Instances. For exanlple, on 5'4-ln. 15.5-lb 5-55 long threads and couplings, the n~etal area (3.299 s q in.) tintes the ntininlunl yield strength (55,000 psi produces a total yleld strength of 181,445 lb. The joint strength under the deslgn factor of 1.4 is 176,- 428 Ib, and this, divided into 181,445 lb, glves a factor of 1.03 on the yield strength. Exceeding the actual yield strength would produce joint failure and this IS runnlng close to the nlininrum. It IS evi- dent that the actual )ield strength has not been ex- ceeded In any of the tes ts , a s no f a~ lu res occurred. I do believe this is an important point to check, however, when d e s l g n q wlth a low tension factor. The risk of exceeding the joint yleld strength is not great, but it is ln~portant when it e x ~ s t s .

Thls paper IS a distinct contribut~on to the knowl- edge of what caslng will actually do. l 'he procedures used In developing the data were posltlve and the results defin~te. It is to be hoped that other con!- panies w ~ l l ~ n ~ t i a t e tes ts of a s in~i lar nature and make further ~nfortnatlon of t l ~ ~ s type available. It is certaln that the importance of reduced design factors IS brought before us here .-In a ntost convincing manner.

Mr. Saye. X l r . Ladd has raised several interesting p o ~ n t s which w ~ l l be discussed tn order.

F'lrst, a suggestion 1s maJe that additional rlsk of caslng collapse would be Incurred under conditions of abnormal external pressures or severe swabb~ng. Fie believe the cllance for such failure is remote if forn~ation pressures have been successfully con- tamed by the mud whose hydrostatic pressure is presun~ed to act on the casing string con~pletely de- void of Ru~d External pressure could, of course, be increased arttficially-for exanjple, during a fracture treatment-lut such conditions would be under the control of the operator.

Second, the question is asked ds to whether niin- illluni yteld strength or ntiniruunl ult~nrate tensile strength 1s used In c a l c u l a t ~ n ~ burst strength of casing. Although our field tes ts of casing-string design factors did not include burst tes ts , tt is well t o s ta te that nllnlniuni internal y ~ e l d pressures are used in Shell designs, these pressures being deter- 1111ned from Uarlow's f o r ~ ~ ~ u l a using ilPI nrinimunl y ~ e l d strength. I'he strength value used In design IS 87'4 ~ e r c e n t of the result obtained by Barlow's for- n~ula , thereby allowing for the API Fbermissible minus tolerance of 12'hpercent on nonlinal wall thickness- es . This value is, of course, that w h ~ c h is tdbulated in API Bul 5C2 and the several caslng handbooks.

l'hlrd, blr. Ladd r a s e s the question of possibly exceeding nilnlrtrunl yield strength under the las t perfect thread when the tension design factor is reduced to 1.40. Nornlally, '4PI j o ~ n t strength, a s de- terrn~ned by enll>lrlcal forniulae In Bul 5C2 and supported by thousands of joint pull-out t e s t s , is used In designing caslng strings for tenslle load. 'Ae belleve that thls is s t i l l the best design criterion. However, Air. Ladd's polnt is fie11 taken and, even though the chances for exceed~ng the nlininlunl yleld strength under the last perfect thread would seen1 rather remote, ~t Leilooves us to cons~der this pos- slbillty when design factors are reduced.

11. G. Texter (Spang-Chalfant Div., The National Supply Co., l'ulsa)(written): 1.0 Ine, this paper is an extremely interesting one a s ~t d iscusses a subject close to niy heart. Throughout my experience a s a pipe engineer i n the oil country, there have been two subjects which 1 have always felt were grossly rn~sunders tood.

The first one concerns the etiect of bad-looking surface defects on the servlce life of drill pipe. I had always suspected that their bark was worse than tllelr bite dnd was sustained in this belief Ly

F I E L D T E S T I N G O F CASING- S T R I N G DESIGN F A C T O R S 37

the results of running strings of defectrve drill prpe by Thompson-CArr, Inc., and by Stanolind Oil and Gas Con~pany. A. W i'honlpson actually proneered the field work, which is con~parable to the work out- lined in this paper.

l 'he second grossly nlisunderstood subject is that of casing safety factors, or, a s the writers of this paper aptly revise it, "design9' factors. I like the latter designation and would urge that al l engineers adopt it for the future a s being lilore descriptrve than 6 6

safety" factor. l'he design factors in use today were far from

being derived scientifically. They just grew. Ap- parently someone, it may have been a manufacturer, once ventured that the design factor (safety factor) in tensron should be a t least 2.5; and s o it was. l3y the same token, the factor for colldpse was agreed upon a s 2.0; both based on average properties, and gradually lowered a s experience seemed to dictate. So far a s 1 know there was never any field testing done on the subject. So, it is now Shell who have pioneered.

I have nothing but praise for both the field work and for the paper rtself. I can s e e nothing wrong with either thelr modus operand1 or with t h e r des- cription thereof, and I am extren~ely pleased that al l the detarls are belng made public I have known many of the Shell engineers who worked on the project , and who often consulted brrefly wrth nie, and they will recall that I consistently urged wrrtrng ~t into an API paper. I feel honored that I was mentioned rn the acknowledgn~ent a s hdving given " a ~ d and advice."

hly criticism of the project, soon after rt was started, was that the conclusions were golng to be based on too few tes ts . I am afraid that this s t i l l holds true from a stat ist ical standpoint. Only 8 tes ts of the tension phase and 5 of the collapse are rather few a s con~pared with the thousands of tes ts involv- ed in arriving at the present API mlninlunl performance values. However, the fact that some 40 wells have since been completed on the b a s ~ s of the new lobered design factors, without trouble, is rather convincing evrdence that the figures are reasonable and practical. I have the feeling that they are.

Of the final conclusions, I am quite intrigued wrth the design factor of 1.4 in tension. In May 1952, answering a foreign operator's query a s to possible nrinimunl factors, I wrote the following a s my personal opinion, " ... I would be satisfied with a safety factor of 1.5 in tension with just ordinary care in

handling, and ... I would go down to 1.4; but I n such case would rnsist on a responsrble engineer belnp present. . " Naturally, ~t is gratifying to have one's guess confirn~ed by actual experin~ents.

Also, a s far back a s 1937, during the drrlling of the early 12,000- to 13,000-ft south Louisiana wells, I belreved it was rather foolish to nlainta~n collapse design factors above 1.0 for the Lotton1 sections of oil strings, and often s o expressed myself. Ilowever, I came to a slightly drl~erent (personal) conclusion from that of the Shell engineers. hly suggestion was to hold a factor of 1.125 down to the last change point (in combination strings) and then let the factor drop down to 1.0, regardless of where the top of the cement might be. Ilere, again, the results of this present investigatron more or less confirnied my early conjectures.

1 nught add one other

3 8 J. E. SAYE AND T. W. G. RICHARDSON

signlng casing strlngs should take Into considera- tlon thls p o s s i b i l ~ t ~ and niust not assume that a particular lot of caslng is representat~ve of the grade, Insofar a s ~ t s particular nlinlnlunl propert~es are concerned. If the material that was used for the field tes ts had propert~es equal to the nllnimunl propert ies for the g a d e , then the tes ts would, it i s believed, have indicated failure in collapse a t a high- er factor of safety than that obtained by the authors, uslng a h~gher than minimum grade product.

In the tension t e s t s the use of a factor of safety of 1.4 will, in some s l zes and we~gh t s of API casing, result In the load approach~ng and even exceeding the y ~ e l d strength of the 111eta1 a t the root of the las t perfect thread. It should be remembered that the API round-thread casing j o ~ n t can be expected to leak when the tension applled to the joint i s equal to or exceeds the yield strength of the joint. The authors are referred to a paper by Thomas and Bartok that was presented at the 1941 API Annual Rleeting which discussed leak-resistance tes ts of casing joints in tension. It IS brought out In that paper that the API round-thread casing jolnt can be expected to be leak-res~stant a t t e n s ~ o n loads up to the y ~ e l d strength of the jolnt but not In excess of thls load. Also, In a discussion of the Thomas-Bartok paper, T. RlcLean Jasper cited the results of tes ts that A. 0. Smith Corporation made of the leak reslstance in tension of the AFI joint which showed that leakage occurred a t only 55.2 percent of the pull-out strength of the joint. There is no infornration glven In the authors' paper a s to whether consideration

. .

was given to the leak reslstance of the API caslng joint when arrivlng a t the factor of safety of 1.4 In t ens~on . It has been and IS now a practice of many operators to use a factor of safety of 1.8 in tenslon, not with the thought of fallure of the jolnt by pull- out but to not load the API short-thread casing joint to the point that leakage would occur.

Mr. Saye: The fact that mininruni yield strengths of the 5-55 and tI-40 material used in the field collapse tes ts compare s o closely with the API average values lends credence to the belief that slnlilar results could have been obtained w ~ t h other API caslng selected a t random. However, we are in full agreement wlth Mr. Uunlop that some nlargln of safety should be provided t o allow use of a joint having absolute minin~un~ properties This is emphasized by the considerable spread in d e s ~ g n factors a t whlch uncelnented casing collapsed. liegardless of whether the absolute value of these design factors a t collapse failure are correct, we believe that the spread in the values does not warrant a reduction in collapse design factor below 1.00 for uncemented caslng. Cy holding d e s ~ g n factors to thls value, an ample nlargin of safety should be obtained to allow for joints havlng API mininrum values. Hundreds of Moody reports have lead us to believe,however, that the chances for approaching this nlinlmunl value are very slight. Rie would be more prone to be concerned over possible damage to casing in t r a n s ~ t , the damage being of such a nature that the tube's r e s ~ s t - ance to collapse would be reduced.

With regard to the etlect of tension on leakage reslstance of a joint, we are fanliliar w ~ t h the work of Thomas and l3artok whlch showed that when ten- s i le load approaches the yield strength under the las t perfect thread the jolnt can be expected to leak. It is ln teres t~ng to note, however, that a l t t ~ o u ~ h well pressures encountered In the Elk City Field made joint leakage an inlportant conslderatlon, reduction of the tension design factor from 1.60 to 1.40 did not result in thread leakage. Prior to using lower t e n s ~ o n design factors, our thread-leakage problem had been -solved, a s by others in the industry, by using good pipe handling, makeup, and running practice.

field testing of casing design factors

Documents