Oilfield ma X ‘Oetober 1991Foc u s The Dolomite Transform Enigma N ulron logging, the measurenient of cap- ‘ture gamma rays or neutrons to determine porosity, hhas a 45-year history checkered with unexpected elfects. One that persisted until recently was a Problem withthe dolomite porosity transform of the ‘compensated neutron tool, which uses two det tors to reduce borehole effects and increase depth of investigation. A porosity transform, or response curve, relates porosity tothe ratio of counting rates ofthe two detectors in dolomite formations. Using the appropriate dolomite transform can make or break the assessment of potential reservoirs. Based on lab data, the or form gave porosities that were too high in the field. Log analysts, following the adage, “the truth is in the field,” rejected the so-called lab dolomite ‘transform in favor of afield dolomite transform. ‘This transtorm was based on core analysis and logs {rom the uncompensated sidewall neutron porosity Buk density, g/om® aol Buk denaty, g/oms ol ld Transforms 0 a Neutron porosily indo, pu 1.038) Denslty porosty,p.u.(metrix= 2.71, fuis 2.71, fuid.098) Density porosity, pu. Data crossplots mado with the field dolomit transform (above) and the new dolomite transtorm (below). Oilfield Review{ool The field dolomite transform, however, unex- pectedly fll shor: it underestimated porosity sub- stantially and could procuce significant erors in thology prediction when neutron and density data were combined in 9 neutron-densiy crossplat Following its widespread acceptance, the eld dolomite transtorm underwent piecemeal mocitica- tions during the 1970s, which included “straighten Ing” at high porosities and “bending” at low porost- ties, The need for improved accuracy and better Interpretation in complex reservoirs led toa major Investigation of transforms and environmental fects, starting in 1980. New transforms developed in 1998 correctly represented the neutron tool response under a variety of conditions. Only slight ‘changes in the od transforms were required for limestone and sandstone, but the new dolomite transform departed significantly from its predeces- sor, Comparison of neutron-density crossplots shows the dramatic diference between the old and new transforms (previous page) A recent evaluation of dolomite reservoirs pro- vides a striking example of the problem and the eco- pact of its solution. The diference in neu- tron porosity estimates produced by the old and new dolomite transforms is obvious (above, right. The reservoir represented inthis log contains ce dolomite wit slight amounts of quartz, residual ol nd fresh invasion water. The lft rack af hel ‘generated withthe old dolomite transform fr con October 1991 Dept, ft (Old Neutron Porosity Doloite Density Persity Dolomite Now Nautron Porosity Dolomite Density Porosity Dolomite = 770 Comparison of old and new dolomite trans! tron and density porosities. rms for neu verting neutron count-rate ratios into porosities, shows a discrepancy between the compensated neu- ‘ron and density logs. “Ola” ‘as much as 7 porosity units (p.u.) lower than those derived from the density log. The right track shows how “new” neutron porosities agree much better. (Enhanced vertical resolution of he right track rosults from alpha processing and is unrelated to use of the new dolomite transform.)— Frosh water — Saltwater Slowing down length and cifusion length, em Porosty +2 pu. ‘The shaded areas indicate where porosity est mates exceed this one's producibilitycutotf of 5% porosity. Although the old transform would have condemned the zone in question, the new one pre- dicts significant production. For the main reservoir, estimates of total thickness increased from 77 ft (23 1m] to 376 115 m] and oll volume from 3.3 hydro- carbon-ftto 23 hydrocarbon ft. Furthermore, using the old transform, GLOBAL and ELAN interpretation programs gave results thal were sensitive to assump- tions about the purity of the rock matrix. Assuming a quart-dolomit misture rather than pute dolomite, the programs predicted higher, but stil incorrect, porosities and quartz concentrations. Using the new transform maintained porosity estimates and cor- rectly predicted small quartz concentrations. Inthe 1970s, dserepancy between lb and field ‘olomite transforms was attributed to absorption of thermal neutrons (which conventional compensated neutron tots detect) by traces of horon and gadotin- lum. Subsequent chemical analyses showed, how- ver, that not enough ofthese trace elements exis din oilfield dolomites to cause the problem. More than a decade passed before the dolomite transform 6 Calculated values of slowing-down length and diffusion length for freshwater lim stone and saturated saltwater limestone. Salinity effect in dolomite, sandstone porosity. - - nigma was corectly explained in terms ofa salini- ty offct and the possible presence of anhydrite. Chlorine n salt water absorbs thermal neutrons, ‘making apparent porosity higher than true porosity. ‘ALthe same time, sodium and chlorine ins push water molecules farther apart, reducing hydrogen concentration; this in turn lowers apparent porsiy. Generally, thermal neuron absorption dominates, How these phenomena influence the formation parameters that ultimately determine the too's ‘count-rate ratio are illustrated for limestone (it) Thermal detection depends on the thermal neutron diffusion length, which decreases significantly as salinity increases. It also depends onthe slowing- down length ofthe neutrons, which increases slight- ly with salinity and with decreasing porosity. The lb dolomite transform was determined with fresh water in borehole and formation. This explains why the lab transform did not perform well under saltwater conditions of the early field trials. Con- versely the field dolomite transform, based in part ‘on data from satwater field conditions, often worked well in saltwater formations, but gave low porosity estimates in fresh water oro ‘The salinity effect in dolomite, imestone and sandstone was modeled (below). The ordinate, expressed in apparent limestone porosity units as a Limostorte ‘A eppaventimestone porosity, pu os ue eS OS ‘Tue lithology porosity, pu,Buk density, g/om® ‘Apparent limestone porosity, pad, ‘normalizing reference transforms are mistakenly used when saturated salt water is present, as was the casein the iil field trials ofthe tool. These errors are simitar, but of ‘opposite sign, o those produced using the field ‘olomite transform when ol or fresh water is present. Independent ofthe salinity factor was anhydrite, which nad been established in some dolomite for- mations by cores, cuttings, photoelectric lags and the sultur yield from the GST tol. Although anby- Arte nasa slightly lower limestone equivalent porosity than that of dolomite, its density is about 1.1 gfeme greater, corresponding to about 7 pu. In ‘the presence of anhydrite impurities, data points shift downward on neutron-density erossplots. When this occurs inthe mid-porosity range, the effect resembles that generated withthe fel transform in freshwater conditions. At very ow porosities, where the water salinity etfect is very smal, the erosspot points respond exclusively to the density increas This sometimes makes a string of density -versus- neutron data are downward as zero porosity is approached (above). These observations strongly suggested that the field dolomite transform suffered from anhydrite impurities as well a om the salin- tyetfect. The truth may bein the fed, but this dolomite transform story teaches us yet again that uncovering itoften takes careful study and cunning —sT Is the error when freshwater Downward shift of data points on a nou- tron-density crossplot at low porosities, indicating an anhy- drite effect. ‘Acknowledgemer ‘and Further o For helpin preparation ofthis ariel, thanks to Dan Arnold, Haliburton Logging Services, Houston, Texas, USA; Darwin Elis, Schhunberger-Doll Research, Ridge field, Connecticut, USA; and Charts Flaum, Schlunber- (er Technical Services, Paris, France. {Inthisarile, GLOBAL (Comprehensive Formation Eval ation Answer Produc), ELAN (Elemental Log Analysis), ‘and GST (Induced Gama Ray Spectrometry Tool ae ‘marks of Seunberger. For the new dolomite porosity mansform: Poses G, Francisque JH, Flaum C, and Wer JP: "The Inpact of Usilzing the New CNE Transforms and Environ ‘mental Corrections in Evaluation of Low Porsity Dolomite Formations,” Transactions ofthe SPWLA sth European Formation Evaluation Symposia, Budapest, Hungary, October 22-26, 1990, paper EE Galford GE, Fann €, Gilchrist WA Jr Soran PD, and Gardner 35: “improved Environmental Corrections for Compensated Neuiron Logs.” SPE Formation Evaluation 8 (1988):371:376. Fr the effects of thermal neuron absorbers Ells DV, Flaun C. Gafort JE and Seou HD: “The Effect ‘of Formanion Absorption an the Thermal Neuron Porosity “Measurement.” paper SPE 16814, presented a the 62nd SPE Annual Technical Conference aad Exhibition, Daas, Texas, USA, September 27-10, 1987 ‘Aruold DM and Smith HD Je: “Experimental Determina tion of Environmental Corvections fr a Dual Spaced Nev tom Porsiy Lag.” Transactions ofthe SPWLA 22nd Annu Logging Symposiam, Mesico City, Mevco, June 23-26, 1981, paper WV. For explanation ofthe salinity and anhydrite effects Bilis DV and Case CR: "CNT-A Dolomite Response, Transactions ofthe SPWLA 24¢h Annual Logging Symp. sinny Calgary, Alberta, Canad, June 27-30, 1983, papers For the origin ofthe fleld dolomite transform Alger RP, Locke S, Nagel WA and Sherman H: “The Dua Spacing Neuron Log CNL," Journal of Pemoleun Teh nology 25 (1972): 1073-1083,State of the Business: Western Canada ‘SASKATCHEWAN Natural gas © Conventional I otcands Deposits of gas, heavy olf and ds in westerh Canada. W... Canada has never beon an place forthe oll business. The geology is compl ‘many targets ae small drilling can be dificult and expensive, about 20% of crudes are heavy and weather significantly limits feld operations tothe coldest six months. The 1990s, however, hold prom- {se for oll producers inthis sacond largest country ‘inthe world, but not without considerable ingenuity. Conventional oll production isin irreversible decline because of maturing ol fields and dwin- ‘ling prospects for tinding replacement light and ‘medium crudes. Abundant natural gas has been {ound and plenty of potential remains—but surplus- ‘ trouble the industry as gas prices stay low. Immense deposits of heavy oll and oll sands assure long-term productivity, but ony for companies large ‘enough to afford the high stakes. To meet these ‘challenges, Canadian companies have made @ ‘commitment to leading edge technology, paticular- ly three-dimensional (3D) seismic surveys, to improve finding costs in region of small and com- plex reservoirs. Despite dificulties, Canada ranks among the top ‘ton oil producers and i third in natural gas produc- tion, Nearly all of Canada’s hydrocarbons come sy Oilfield Reviewfrom the Western basin, which spans 2500 km [1500 ‘miles] from Manitoba, beneath the prairies of Saskatchewan and Alberta, tothe foothills of the Rockies in Alberta and British Columbia (previous age), By far, most oil and gas reservoirs lie in Alberta, which contributes 80 to 90% ofthe coun- tty's hydrocarbon production. For this reason and historical precedence, the Canadian energy industry is concentrated in Alberta's two largest cities, just ‘300 km (180 miles) apart, Calgary (population 708,000) and Edmonton (population 605,000). Cal- gary is home to several hundred oil companies, most within easy walking distance of each other. Edmon- {ton is surrounded by oilfields, having more than 2000 producing welts within a 40-km (24-mile) radius. Like its counterparts worldwide, the Canadian ‘energy industry was hit hard by the collapse of oll prices in 1986. Since then, industry employment hhas dropped 40%, a loss of 50,000 jobs, and the industry average return on capital has fallen to @ meager 4%, below the for investors (above, right). “We didn’t think it was possible to live with ‘$15/barrel oil,” said Doug Stoneman, executive vice President of Shell Canada Limited. “But there is life at $15. Today, the industry is slimmer, more {focused and more efficient.” | considered attractive October 1991 % T Return on capital for the ol industry in mi western Canada. #10 3 E 4 = 2 | a a a a) Year Oil companies survived by rationalizing assets ‘nd restructuring, selling or trading interests in properties that contributed too litle profit. in recent yeats, billions of dollars have been paid for oll and gas reserves that were on the market for less than their equivalent finding and development costs Despite improvements inefficiency, the Canadian industry faces further challenges, mainly because of aging of the Western basin. As fields mature, pro- ‘duction dectines and more wells are needed toreplace ost production. Yt, overhe past decade in Number of wolls and Alberia, average paucity of conventional oi OY Mell productivity well as dropped more than 50%, while the num- ao ber of operating wells has grown 80%—not enough g to keep pace. As crue production has fallen, oper- fe 5 ating cost have sien rm 20% to neat 40% of a j oross revenue. “We're at the point in Canada where it's tough to 20) find new oil reserves,” said Gwyn Morgan, presi- J dent of AEC Oil and Gas, a division of Alberta Energy ee We ae ee Company. “Next year, western Canadian reserves ‘eer wil dectne despite maintaining production. No one — Arora wetsin operation ‘doubts that we're on the down side and that we i Bareis per welt por dey not going to reverse he slope. This means any: sonable ol prospect wil be drilled within any rea and easter Alberta il sand reserves are estimat- sonable range of ol prices" (ight) ed at 1.7 tllon bares, more than double the con- Immediate recovery of Canaia's energy business ‘ventional reserves in he res ofthe word. The ‘may not rely on the country’s immense deposits of 307 billion barrels that may be recovered economi- oil sands (known colloquially as tar sands) and heavy Cally equal nearly half the world's known reserves. oil, but they play a roe in its long-term prosper. With so much heavy ol underfoot, the Canatians These deposits are concentrated mainly in northern have become resourceful al ining ways to produce it, even at todays prices. At he present production af more than 220,000 barels per day, ungraded light crudes from heavy oll account for 20% of Cana- da's light and medium ol supply. Some heavy oll canbe transported by pipeline with the ad of duets. Some, inthe form of bitumen- saturated sands is mined and upgraded near the source. Ths extraction is perfrmed inthe Athabas- a oil sands northeast of Edmonton, which were first noted by fur traders inthe late 1700s. Government- © sponsored studies of the oil sands began in the 1800s, and oil was first extracted in the 1920s using hot ‘water flotation, a process still used today. Seven Canadian ail companies andthe govern- ‘ment own and operate the Syncrude projet, which Drag fine mining of extract ll rom mined oll sands Huge draglines oll sand by Synerude — segop yp the oil sand and deposit the mined materi- Soe geen’ al in kilometer-long heaps (left). Bucket wheel McMurray, Alberta. For seale, a truck Is 10 Oilfield Reviewreclaimers collect oil sand from the heaps and deposit it on conveyors that carry up to 6300 metric {ons {7000 tons) per hour to an extraction plant. In ‘the extraction plant, ol sand is dumped into large ‘tumblers and mixed with hot water and caustic soda, which begins separation of bitumen from sand. Treatment in settling tanks, followed by cen- {rifuging, futher separates bitumen from clean sand. Bitumen then goes to upgraders for conver- sion toa light synthetic erude oil. Primary upgrad- {ng converts bitumen into liquid hydrocarbons by ‘racking reactions at high temperature with coke ‘and steam. Secondary upgrading remaves impurities, mainly sulfur and nitrogen, leaving purified naph- tha, gas oll and a small amount of butane. These ‘components are mixed to form a light synthetic crude that can be handled by conventional refineries. Oil companies have a healthy respect forthe com- Plexities of Canadian exploration and endorse ‘emerging technologies as a means of lowering find- ing costs. Their use of 2D seismic technology, in particular, Is often cited as a leading factor in thelr suecess in Canada, ‘There's no question that 3D seismic has been key to our success in both deep Devonian and foothills gas plays,” said Guyn Morgan of Alberta Energy. “We spare no effort because land and Arilling costs are so high. We also use 30 seismic when we're trying to drill small targets at high cost, 0 We use iin almost all of our oll exploration because we're looking for small reefs and, in some cases, small structural sand plays.” October 1991 Inthe late 1980s, Alberta Energy reported a show in deep Devonian well. A few years later, with 3D seismic and better interpretation, the company ‘determined the well was on the edge ofa thick reet. “We drilled a new well clase tothe old one,” sald Morgan,” and came right into the reel. That's what {3D does for us.” ‘This entrepreneurial sprit has carried much of the Canadian petroleum business through hard times. But more than gusto will be needed to litt western Canada’s gas producers out of the doldrums. They are caught in a supply-demand vise. They have abundant long-term supplies with much more to be found. They supply about 7% of the gas consumed by their neighbor, the United States, the largest energy market inthe world, which concluded a con- troversial Free Trade agreement with Canada in 1988. What's more, gas is the choice to meet clean- nmental restrictions, air standards and satisty em and! per unit of energy is about half the cast of oil. Yet, two thirds of the annual 100 billion eubie ‘meters (3.5 trilion cubic feet) of production is for household consumption, which is stagnant, whi {ndustrial-commercial consumption is decreasing, This leaves export tothe US, which last year hit a istoric high, growing nearly 4% over the previous, year. But domestic declines lowered overall sales. u200 1.50) 1.00 U.S. dolars per cube 50 a Year Gas prices eroded by more than 40% through the tate 805 tothe present, brought on, many believe, by a sudden change from gas price regulation to deregulation in Canada and the US (above). The gas ‘shortages ofthe 1970s, resulting from price con- trols, forced the majority of industrial gas users to install dual fuel capability. After deregulation in the fate 1970s, gas prices rose toward parity with oll, setting off a boom in gas exploration. Users responded to price increases by switching ta oll and coal, leaving a glut of gas, the so-called gas bubble that stil hangs over the market today, This surplus has been maintained by mild winters, naw US gas Aiscoveries and increasing exploitation of technolo- 4y for producing methane from US coal beds, a potentially huge reserve. ‘Gwyn Morgan of Alberta Energy foresees contin- uod hard times for Canadian gas producers, atleast inthe short term. “There stil are significant gas reserves tobe discovered and the Western basin is mainly a natural gas basin, But activity is going to be driven by prices and markets, and that will mean les activity.” New markets for Canadian gas should open when $5 billion in pipeline expansions are completed over the next few years. This will increase capacity ofthe Gas prices over the past 17 years. ‘TransCanada Pipeline and add new pipelines to ‘southern California and the New York metropolitan, area. U.S. electricity demand is growing, although ‘capacity is adequate thanks to a buildup in the 1970s. Should new generating capacity be added, utilities undoubtedly wit! favor combined-cycle plants, which harness waste heat from turbines to ower other turbines. These plants have operating costs nearly 30% below those of oil- or coal-fired plants, and provide some environmental benefits. All this bodes well for Canadian gas companies in the 90s. SM i F Ere ‘Ayoub J, Colson L, Minke J, Jolmsion D and Levine J ‘Learning to Produce Coathed Methane," Oilfield Review 3,na,1 Janvary 1991): 27-40. General reading on the Westera Canada petroleum industry The Syncrude Project. Calgary, Albert: PanCanadian Peolewn Limited ‘Nikle's OF and Gas States Quarterly, Third Quarter 1990, Calgary, Alberta: Southam Business Information ‘and Communications Group In, 1990. Hayes TC: “Bowom-Fishing inthe Gas Patch,” The New York Times, May 19,1991, section 3. 1990 Annual Report. Calgary, Alberta: Canadian Petro leu Association, 1991, For help in preparaviom ofthis focus, thanks to the follow ing people hased in Calgary, Alberta, Canada: Donald Barkwell, Paco Petrofeun Lid; Ron Hicala, Canaan: Hunier Exploration, Lad; Myron Kank, Alberta Miniry of Energy. Gwyn Morgan, Alberta Energy Company td. Bartlert Rombough, PanCanadian Peroteum Linited: lan R. Smyth, Canadian Petroleum Association; D.G, Stone ‘man, Shell Canada: Guy J.Turcone, Chaurco Resources, Lid; Joe Amin, Bud Bell and Lue Laverdiere, Scan: berger of Canada, Oilfield Review© Causes, Detection and Prevention It costs the oil industry between $200 and $500 million each year, occurs in 15% of wells, and in many cases is preventable. Stuck pipe remains a major headache that demands and Is getting industry-wide attention. Louise Bailey Tim Jones, Cambridge, England Jim Belaskie Jacques Orban Mike Sheppard Sugar Land, Texas, USA to Houwen Stuart Jardine Dominic MeCann ‘Montrouge, France Forhelp mn peparsan oT i anil, ovis whe tre tah Peau (8 suck pipe ek orc Rab f TOTAL ones’ Balipapan J Steve Devin, Pl Hammad, Gey fe Shera, Scene Care i Cambie Ea ‘The following fictionalized conversation describes events that actually occurred on a tig driting offshore io the Gulf of Mexico. It is noon on a Monday and drilling is at 3470 feet. The mud logger has just joined the ‘riler onthe rg or, October 1991 Mud Logger: We've drilled three fer into a break, Better pick up and check {for flow Driller: One flow check coming up. ‘Aflow check tells whether a kick isin ‘progress. This is routinely performed after a ‘rlling broak, a sudden inerease in penet tion rate, usually indicative of permeable for ‘mation. Tho mud jogger returns to his unit, A shart time later, the driller phones. Driller: We've shut down for five min- ies and there hasn't been any flow. We're going back to drilling, Mud Logs Before recommencing diling, the driller picks up a lite to work the plpe—and then calls the toolpusher. Driller: The hook load shot way up when I picked up after the flow check, We must have had! 100 kib of overpull ‘for a couple of seconds there. Toolpushor: How does it look now? Driller: Everything fine. Toolpusher: Keep an eve on it. 'll go talk to the company man. BFitleen minutes ater, the measurement- hile-diting (MWD) engineer appears on the deil floor MWD Engineer: Did you notice the overpull when you picked up? Driller: Sure did. It was over 100 ktb. What are we drilling down there? MWD Engineer: The gamma ray shows a clean sand, and it looks like there night be something in it too. Thirty minutes late, inthe mud logging unit: ‘Mud Logger: How's the mud doing? ‘Mud Engineer: Lousy. The nuul cleaner and desander take turns breaking down The company man is upset about the high water loss, and the nuid weight keeps increasing. I'm stuck between a vock and a hard place. Mud Logger: I ber te nuudstone we drilled earlier is feeding the mud sys- tem. That stuff was soft—it practically dissolved in water. Mud Engineers Phat must be it, The drill solids are steadily increasing. It ‘must be coming from the open hole somewhere. High water loss and inereasing dil solids produce thick, sof mudcake, a primary cause of ditferenial sticking. Fifteen minutes tater, atthe mud pts Company Man: When are we going to reduce this water loss? Mud Engineer: I'm trying everything 1 know, but the mud cleaner is down and that mudstone we drilled is putting a lot of solids into the system. We need to get it fixed and the desander needs work too. Company Man: Tlie rental company says they won't have a man out for another day at the earliest. We'll just have to make do, at least until the crew ‘change tomorrow morning. 4 (One hour late, inthe company man's office: MWD Enginooe: Pore pressure is still «at 87 ppg. Is there any reason the mud weight has been increased to 94 ppe? Company Man: Yeah, we can't keep the solids control equipment running The weight is creeping up. MWD Engineer: Wien wre shut down {for the flow check in that sand, we had 100-kib overpull. We could be geuting some differential sticking. Company Man: Maybe you're right. P'lcall the drill floor and have them Keep the pipe moving. Let me know when we get out of the sand, ‘The next morning, the mud cleaning equip ‘meat i stil malfunctioning, but drilling pro- ceeds. Meanwhile a helicopter arives with a ‘new company man. Arriving Company Man: ITow's every thing going? Leaving Company Man: Pretty good. But yesterday, we had a 100-klb over- pull after a flow check. We haven't seen anything since. Arriving Company Man: What do you think it was? Leaving Company Man: think if was differential, We'd been having problems with the mud cleaners. Everything is in the reports. I'd beter get going, the heli- copter's waiting. See you in two weeks. Arriving Company Man: Have a good time off Later that evening, total depth (TO) is reached. The driller cals the new company man. Driltor: We just reached TD and are going to pull out. You want me to rotate when breaking connections? Company Man: No, let's not waste any time, We've got a long logeing pro- gram and we're behind schedule. Driller: OK. Oilfield ReviewThey pull five stands and then cannot raise the pipe aera connection. The bottombole assembly (BHA) happens to be opposite the Sand where the dling break occured and has became diferentally stick pressured agaist the permeable sand so hard that no force on earth can move it. Iisa drillers nightmare that is less common than ten years ago, but till common enough. In most cases, petting stuck is preventable and the _main reason it happens is hat warning signs come sporadically over days or even weeks. This pipe got stuck because warning signs were poorly communicated and eventually forgotten. If the mud cleaning equipment had been working if the pipe had been kept moving while pulling out, ithe ig had had the advantage ofthe latest information sys tem technology, then perhaps the incident would have been averted. Poor communi «ation i often the main culprit behind stuck pipe, with the many players-—itom com: pany man to shaker hand—not pooling their ‘observations and fling to arrive at a cllec- tive decision atthe right moment. The industy’s interes in tuck pipe is eur rently going through a renaissance. Several operators are making detemined efforts to codify the warning signs and to improve communication—not just for their own aller, but for all on-site dling and service company personnel. Meanwhile, beter rig Sensors an information systems are provi ing sigefloor “Smart” alarms to help the diler recognize touble before it gts out of hand. ‘The causes of stuck pipe have been known since drilling began (next page). Broadly, they are divided among dirential sticking, formation-related sticking and mechanical sticking." Diferential sticking ‘occurs in permeable zones wien dil col lars, drillpipe or casing get embedded in ‘mudcake and pinned tothe borehole wall by the diffrence between the muds hydro- static pressure and a lower formation pres- sure, Formation-related sticking. occurs hen unstable formation constrict the dil string, This includes unconsolidated rock, swelling shale, flowing formations such as salt and plastic shale, and geopressured for mations. Mechanical sticking covers numer ‘us causes such as key seating-—in which a groove cut in the borehole wall by dillpipe October 1991 > (Cross section ofa dill collar (etl) embedded in mudcake and pplnned fo the borehole by the differential pressure between ‘mud and formation, The pinning force is ihe pressure dlference fimes the contact area between ail collar and mucieake. ‘According the Coulomb iriction model, the overpull required to ‘ree the drlstring is the pinning force times the friction facter, which is generally 0.3 0.0.4 1 the dillistring remains stationany for any reason, more mud ‘cake can build up, increasing the contact area and making the dillistring harder to tree. traps the larger-diameter hardware when tripping out—accumulation of cuttings due to poor hole cleaning, undergauge hole, doglegs, junk, collapsed casing, and frag mented cement. The causes of sticking are more numerous than can be listed here, and itis a mistake to think that only one cause may be acting Pipe stuck because of swelling shale may sooner or later also get stuck differentially at another point inthe well. One of the results Of the industry's current attention is a better understanding of the events leading up to stuck pipe and thoir interpretation in terms of the causes of sticking. Knowing the causes is essential for taking correct rere: dial action, Dilferential sticking Research into dferentally stuck pipe started in the 195052 Simplisticaly, the pressure dif ferential between the mud and formation pushes the dil collars and dillpipe against the borehole wall, and the friction force required to move the string—the pinning force multiplied by a coefficient of tric: tion—becomes too great for the rig draw- ‘works fo pull the pipe free (above). In the simplistic picture, pinning force equals the pressure cliferential multiplied by the con- tact area hetween drill collar and borehole (continuon page 18) inthis aie SPN icing Pipe nico ea natok ‘Ani MOS ea marc Seeo Fore 1. Mechanical shi omens ude 0 Ice oat ae lig 2. Oumans HD:* Mecha of Dire Pesre Siting of Onl Colas,” Pevoleun tans, ‘AIMED (1956 268-274 ‘Ani Mand Monaghan Ps -Diecetal Peste Sckng Laboratory Studies of con ewer et and Fite Cae" fouaf of Peelcun fecha! ‘ny 4 19625 3375343. 15reer rr Unconsolidated formations Uneonsoldated for: ‘mations such es loosely compacted ‘sands and gravais ‘an collapsa into tho wokbore forming @ brdge around the sting Fractired/Fauited formation Formation thats paturally fissured lornear a fault zone ‘may break ofl in lees ino te bor. Foland jam tho ir 4 ailisting. Feces can vay rom smal Lp to boulder size, Formations that Commonly recture SS) are carbonates tnd shales, reactive formations Carian monte: lone and bentonite shales hydrate and ‘wal on contact wih Water-base mud, fil i ing the borehole and creating clay bal that oan block the wellbore and consti tha die string. Treating the ud with KCI and polymer can arest hydration Oltbase mud inhibits the pro 68s completoy Mobile formations | Geopressured formations: Formations Ike eat ‘and plastc shales It erally flow iio the = welbore wren recitsinng strosses | |e {are removed, jam: = ining the dillstring thatean slick the PTS — overpressurea > f 1 = | rormations can blow =f] = epanvten pene: trated by the dl bt, fing the borohote wn rock partcles aiilsting 3 i = Oilfield ReviewPoor hole Cleaning Key Seating > Koy coats, grooves in the borehole wal ut by rotating dlpipe, ‘lk larger diameter collars and othor | hharanare wnen tip ping out A key seat Team placed higher in the eting ean opon tho kay sat, allowing | passage of cols. October 1991 Poor hale cleaning reeults in overloading the annulus vith cute tings, potently stick: ing the alsin. ‘Ths is most kel in \Washouts where annu- Tar velocty decreases: ‘and cutngs accumu: lato In deviated wale, ‘outings form beds on ‘the lon side of the hole and can migrate uphole Ike shifting sand dunes, Collapsed casing Collapsed easing cours when forma: tion forces excond casing colapse pres 9, such as whon + casing is too light uty ‘casing is 01d + casings landed vith f90 much ton: son, redulng i callapse rating: unk Junk is eny object in th ole not meant lobe thera Roller ‘cones or aven a POC cutter aro large ‘onough fo sick the ‘sing, Once tho dlsting becomes tee, junk must be ‘ished wih a revorse- ‘culatlon junk bas: ot or magnets, Hl Coment related Cermentiebied sticking oocurs when blocks of coment fall Ino the wellboro from casing ratholes forcement plugs, Jamming rising ‘also occurs when sillsiing becomes planted in soft or *oreen’ cement that flash sets when pressure Is epalod Dorota ising cass bepouia pr | Pee ‘when drill collars get beddedin het ed fuses ius odo ts bore: Fowl oy he (ec pee serves nd ae aan (Natori oripled te BP Beebe ask orcwall, an area that inereases with mudeake thickness and length of BHA in contact with the formation, ‘According to this picture, preventing dl ferential sticking depends on, among other things, careful mud design and condltion- ing. The mud must not be too heavy bbecause that increases hydrostatic head and differential pressure. A recent survey by Chevron USA Inc. indicates the increasing probability of differential sticking with pres- sure and suggests @ maximum of 2000 psi, although this obviously is subject to local conditions (befow). The mud must also hhave good fluid foss properties to prevent excessive mudeake builkup. And the solids ‘control equipment must be functioning cor- rectly to prevent solids ancl cuttings from remaining in the mud and possibly accret- ing o the cake, building it up further. Other preventive measures include always keeping the string in motion, partic- ularly when adding pipe while drilling, the ‘Stuck pipe probabilty, % 1000 Differential pressure, psi 1800 2000 2500 3000 Probability of differentia! sticking versus differential pressure, established from ‘analysis of 600 well histortes in the Gulf of ‘Moxico. An ‘accoptablo" probability of 20% is achieved if diiferential pressure ever exceeds 2000 pal (From Weaikley, reference 3) 18 Ing cake, but ft takes the most time. ‘most likely moment for differential sticking Wiper trips and teaming also help because they remove parts of the mudcake (above). Top drives, which permit rotation while ras ing pipe, contribute to decreasing the chances of getting stuck. Preventive measures, though, must begin in the planning stages. The first step is selecting casing points. These are usually picked so that in each openhole section mud weight lies between a lower limit, below which the mud would no longer ‘counterbalance pore pressure, and an upper limit, above which the mud would fracture ‘or damage the formation causing lost circu: lation. a maximum dliferential pressure is imposed to reduce the chances of ciferen- lial sticking, for example opposite perme- able depleted formations, the upper limit may have to be reduced and an additional ‘casing point may be necessary. The extra cost of adding a casing string, however, must be weighed against the alternative of simply recucing mud weight and hoping the well will not kick—stuck pipe could be less expensive than a blowout. Further well planning is necessary for dviated wells. The well trajectory should bbe designed to minimize drag. Use of steer able assemblies without rotation to steer the hole should be minimized—rotary drilling produces fewer doglegs. {A second planning step is careful BHA design—itis the lage diameter drill collars that usually get stuck. The tend toward sim- plex BHAs, with only as many collars as are strictly necessary, has done much to reduce the incidence of sticking. Further reductions have resulted from using spiral collars, Which present less surface area tothe bore- hole wall, and stabilizers every second or third joint that keep the entire BHA away ffom the borehole wal A thitd way to avoid differential sticking is to use MWD measurements to provide a Continuous record of direction and inclina- tion. Obtaining these data the traditional way, by lowering instruments down the Aillpipe, keeps the drillstring stationary for {Up to 30 minutes, time enough to get tly stuck. Fourth, spotting fluids circulated to free clferentially stuck pipe should be pre- mixed and avaiable at all times. ‘Advances in preventing differential stick: ing currently rely on further research into the sticking mechanism, better undestand- ing and coxlication ofthe telltale signs that precede sticking, improved monitoring of fig data through drilling information and alarm systems to detect those signs, and, ‘most important ofall, training of rig floor personnel that emphasizes the importance (of communication, Researchers seek answers to the most basic questions. What exactly isthe sticking ‘mechanism Is it the Coulomb friction force mentioned above, or is sticking not associ- ated with the borehole wall but entirely with the mudcake? Perhaps the force needed to free the BHA must simply over: ‘come the yield stress of the cake, Pursuing this line of thought researchers are invest tating in detail the properties of mudcakes next page, below leds ‘Ata larger scale, several groups of experi- enters have built wellbore similotos that offer insight into the friction forces that develop while diling or sticking. None, though, has satisfactorily determined what factors contribute to the suddenness or intensity ofthe sicking pipe phenomenon S ‘One more enigma for researchers is why spoating fluids ae oflen successful at freeing pipe. Does the diesel-base fluid infiltrate Ollie Review(iene ‘Flow atagram for dlagnosing the causo of stuck pipe when encountering overpull after moving pipe rom a static position—for example, after making or breaking connections during dilling, tipping or reaming. ‘This and other similar charts were compiled by the BP stuck pipe task force. Prevention of stuck pipe depends on speedily answer- ing each query as it arises, This is boing facilitated by increasingly sophisticated data provided by now MWD tools and better datr analysis and alarms provided by computerized rig instrumentation. (Couresy cf RP Fxpleration) 6 Distance trom titer paper, mm ‘ore analyzing the implications of this and thor results for cifferential sticking. (rom Sherwood et a, Journal of Chemical Sosiety Faraly Tiemsaciions, reference 4.) October 1997 and destroy the cake? Or dos it lubricate the drillstring, allowing moze pulling force to teach the stuck BHA? If scientists can answer these questions, better methods for fteeing stuck pipe may become available. ‘Meanwhile, the emphasis is on preven- jon, And thal means svatching the warning signs and knowing what combination of cir- cumstances presages a stuck pipe incident (One of the aids produced by a recent BP stuck pipe task force is flow diagram for each dailling operation—deilling, making connections, tipping in and out, reaming in and out, circulating and running casing, The diagrams show how events accumulate 10 make sticking increasingly probable.6 Dif- ferential sticking usually occurs when mov- ing from a static postion. The relevant flow diagram leads inexorably through the events, queries and answers that culminate in stuck pipe (above): * Does drag increase when moving string from static position? Yes. «Is circulation restricted? No, + Are permeable formations exposed? Yes. + Can dillsting be moved? No! The BP chart not only forewars of impend- ing disaster, but also suggests what type of, slicking should be expected. For example, if, aiter moving pipe from a static position, circulation had been restricted, we are led to the left part ofthe diagram toward forma- Sy kakey RRs “Use of ick Pipe Sais Radice {heOcraenceet Stuck Pipe pape SPE 20110, pe “tied atthe th SP Anal lec Caneenee [nl shen, New Oreo, Leising, USA. Scptembe 2326 1990 NewoueCC:“SuccetallyDalling Severe Depleted Sands paper SPEAADI 21913, presented tthe 1991 SPEAADC Dailing Coden, Ase. ‘dame The Nee Math Fe, 15 Shen, Meten CH, Far CA an Alera sConcenton Pot Win Novo Ma Sees” uma of Camis Sac Fad ame ters 8990) 61518, ‘Alder NI Mate GH at Shenae Rhcoraryo tert Cal" Jura ef Naw Neto on Fu suchansce 39 (1990 291-310 ‘heron JO, Metan GH, Fae CA ard Aleman ‘Ne"Sgucez'n Rony o Now uterm Mo ‘kes hua of Nar Netenian Fad Mecans Spusen sae ‘Couns and Zuo “A New Approach Oi feel Sk” paper SPE, preset he [anSPE Anal tehncal Confrence sad Eb ton, as Wgas Nevada USA, Sep 2225, te, Quiles, Oslo AX and Zanera MEA Ful ‘Ste Were ton Serle” paper ADCAPE 19858 pst at i DIAS Dlg Co ere Ht, Tt USA Febuary 27-Math 2, "00 ‘5. Bradey WV, rman O, Pst RS, Wood RO Schofield Te Aufl ond Cotkng DATs Force ‘Aeproac a Reducing Suck Pip Coss” pape SSPEIRDC 71995, preside 1991 SPEAADC Dlg Coferene Anstadam, Th Neerland, Nive 1-1, 199 “Flaw Chats cp Prevent Sk Pipe Problem,” Oi ‘Gena 8 nn THA 1, 1991 6b Team it feral in Costing Suck ge, OF ‘8a 99, 9.15 ge 15.991: 3955 19Je section diled 5 previous bi DA mere complex diagnostic flow dlagram for trpping out. (Courtesy of P Exploration) 20 Oilfield Reviewtion-telated sticking. A more complex chart provides an analysis of what may happen while tipping out (previous page). Possible sticking mechanisms include formation- related problems, key seating, cement relate problems, junk and wellbore geometry Shale Sticking Differential sticking accounts for more than 70% of stuck pipe in areas where wells are drilled through depleted, highly permeable sands asin the Gulf of Mexico. But in other areas, such as the North Sea, where opera tors are obliged to dill through unstable Ter tiay shales to reach pay, shale-related stick ing claims a similarly high percentage. In the presence of waler-base mud, shales frequently swell—partcularly those contain- ing abundant montmorillonite—constitin the borehole and finally gripping the drill: string. Others slough off small particles that {ill the borehole and risk packing off the BHA or bit. Drillers are awate of shale- related drilling problems if they have encountered them in nearby wells, and they have several methods of dealing with them, The simplest is to weight up the mud, although this does nat necessaily halt the chemical reactions berween mud and shale that are responsible for shale’s instability The simplest method of contiolling chem ically active shale is to use oll-base mud, a solution that ansformed exploitation in the 'Nonth Sea, By containing its water phase in small eil droplets, oib-base mud eliminates contact between the water and the forma. lion, ensuring that the shale remains stable. In a wellbore simulator at Schlumberger ‘Cambridge Research, Cambridge, England, oik-base mud was flowed through a one inch borehole drilled in a block of Pierre shale—a highly unstable rock obtained irom ‘2 near-surface site in Colorado, USA con: taining 30% clay, of which a third is mont ‘morillonite. In the simulator, overburden, confining and mud pressures can be inde- pendently maintained up to 30 megaPascals (MPa) [4350 psil (top, right). Throughout the experiment, the borehole remained on gauge and relatively smooth, indicating that asin real life, o-base mud eliminates shale- related sticking problems (right)? {As environmental restrictions tighten, though, the toxicity of oil-base mud has October 1991 ‘Mud prassure contol -Contning pressure ‘Shale sample Mud reservoir Sampling vale ‘Schematic of borehole simulator deve!- ‘pod at Schlumberger Cambridge Resoarch. ‘is siraulates mud clreulation through a IVinch borehole drilled in a rock sample ‘measuring 6 inches in diameter and 8 Inches high. The thin mandrl inside the Borehole carries a caliper that can mea. sure the borehole during circulation, Sliced view of a bored core of unstable Plone shale that hes boon circulated with ‘oll-base miid in the borehole simulator at ‘Schlumberger Cambridge Research, Bore- hole diameter is one Inch The borehole wall is mainly smooth, Indicating the lack of interaction between {ermattion andl muc. Tho sometimes ragged fabric atthe face occurs whan the ‘core fs sliced. thrown its future into question. This is spurting inereased research on mutls made with vegetable rather than petroleum oil oF with polymers synthesized from petrochem- Teals. The latter, however, may be just as toxic as conventional oil-base mud. That leaves water-base mud The water-base solution for fighting unruly shales is generally a mix of potassium chlo- ride (KCI) and polymer. The role of the polymer is poorly understood. Current joint esearch at Schlumberger Cambridge Research and Institut Francais du Pétrole, Rueil-Malmaison, France suggests that it plays no cole in arresting the chemical reac: tions that occurs when shale meets water. Rather, its roe isto prevent dispersion by physically encapsulating clay particles as the shale swells and disintegrates The role of KCI is better understood. The common denominator between mud and shale is water activity, a measure ofthe abil- ity of water molecules to escape from a sub- stance. By definition, pure water has a water activity of 1, Anything that gives up water ‘vapor less easily, and shales usually it this category, has an activity less than 1, When ‘vo Substances are in contact, an equalizing ‘reterwelbore siulator ned eames ‘ted peblors becca ‘Simpin ,Deahg Hand Sas DP: “Down ok Senso Col hows Unespel ecteck Sheraton on Boreas Wl paper ADCAPE 17202, eset atthe 1088 ADCAPE 21process takes place in which the substance ‘vith the higher water activity loses water ‘molecules to the substance with the lower ‘activity, Thus, watecbase mud tends to lose ‘water to shales. In absorbing water, shales swell or break up. To prevent this, a water activity balance must be achieved by 1) reducing water activity of the mud and/or 2) increasing activity of the shale. Mixing salts like KCl, sodium chloride (NaCl) or caleium chloride (CaCI, into the mud achieves the first gol, with KCI providing the least reduction and CaCI, providing the most. The second goal may be achieved when mud contacts the shale and the salt enters the shale with the absorbed water. As this occurs, the salt cation—potassium, sodium or calcium: replaces the original cations in the shal, which are predominantly sodium, Labora tory measurements have shown that if cal- cium replaces sodium, the shale activity decreases—a step in the wrong direction Sodium cations replacing sodium-rich shale tive practically no change, which does not help. But when potassium replaces sodium, activity increases. Thus KCI mud achieves the dual goal of decreasing mud activity and increasing shale activity Several laboratory techniques are rou tinely used to measure the activity of shales, shale swelling and the effectiveness of mud formulations in stabilizing shales A simple relative humidity meter measures activity Swelling is measured by compressing 3 paste of shale into a receptacle with a pis- ton, exposing one end of the paste to the ‘mud, and then measuring the outward dis placement of the piston versus time once the compression pressure has been released, The failure or success ofa particular mud in controlling shale dispersion is determined in 2 hot rolling oven. Small particles of shale are weighed, then placed with the mud! in \Btect of water-baso muds on the Piorre shale. A KCI mud! with activity balancing the ‘shale’s Cot produces a relatively smooth borehole wall. An unbalanced mud stimu: Jates considerable hydration and creates c rough wall (ight). In the wellbore, shale ‘patticies would have fallen on the BHA and bit, possibly packing them off closed containers. The containers are rolled gently overnight in an oven, to enhance the mud:shale interaction at downhole temper ature conditions, The next morning, the resulting soup is sieved to retrieve remain- ing particles, which are then dried and ‘weighed, The goal is to have minimum loss of shale into the mud. A good inhibitive KCl-polymer mud will typically preserve 95 to 100% of the original shale. Using larger ‘equipment such as the Schlumberger Cam- bridge Research small wellbore simulatot, the effect of incorrect mud formulations can be directly observed; the results are often frightening (above) ‘Measuring shale activity in situ is @ much harder challenge. Currently, the only hope is from geochemical logging measurements that can be interpreted to provide a sophisti cated mineralogical analysis of the forma. tion. Ifthe activities of each mineral compo: nent are known, the analysis may then provide an estimate of formation activity. Logs either come after the well i drilled and the risk of sticking has passed, or from nearby wells. Ideally, the same technique ‘would be used in real time with MWD data, Curent MWD technology, however, fails © provide the necessary range of geochemical Real-time Prevention In other ways, MWD technology provides valuable data for preventing stuck pipe in real time. in the diling made, MWD mea: surements of downhole weight-on-bit and downhole torque can be compared with surface hook load and torque. Using a model of the well trajectory, also deter ‘mined from MWD measurements, the com: parison permits an estimate of pipe friction versus depth that can provide early warning, Of sticking.® Anadiill’s SPIN sticking pipe indicator works on this principle and pro- vides both dag and rotating fiction factors (next page) A recently introduced MWD wellbore caliper measurement provides another indi cation of sticking by directly showing how mobile formations and unstable shales affect wellbore diameter (p2ge 24).° The evolu: tion of welloore diameter can be monitored during each trip in and out of the hole. This new sensor promises easy identification of key seats Pipe friction in the tripping mode is one ‘output ofa trip monitoring program recently Oilfield Reviewsno] Gamma rey, | Surface Surleoo — | Drag aon ountesee | weights | § | _ | torque, kb | factor, jo ——eo] 040] @ | = [o—_ 20 ‘0 Interpretation Faeot | ows | 8|€| pron poneraton, | dowmroe | = | & | cowntole | Rotary tiction tw | wert | & | | torque,rtia | fact. % ee Qe fo ——— 0 ax OP increaos to 70th Fotary Woon decreases ony a wien Gable ave san 24+ | Dlagnosttemedate sa orshangng in sondat 310 = rag and ity cton Inerease: HOP decrees tom S010 15 try Wong ane Ea feornng ol as no sucess. Fiveslond viper wo decreases bon cts noma ves. co ROP roreaes to coun Dianoss ered tabi ter ggg sand at ro Rotary ton increas sharp rc chaginerensae soy, ROP vo Secteses fom 61010 Three sind wpe ip creases bot dag and lary ave Ht ROP eas 60 Diagost top saber hang ingin cand 710 | | cop $ ‘et «ccanecon, holo was Fe orton and reamed Roly Ho: ton dcreasos chap vod Diag freon incense. Tree sand wiper trip brings it down Diagnosis: top stabiizer hang ing i sand at 80 Naidu A aed Calanibor Designo Cheracally alone Pome ring Fld Geto s Rehtton in ly Delran” YE ring gern 14198603112. ‘Geneve ME and Ose SO, “hae ih tion Dare th Eecring4 1999 261268. 9. erage Malco’ G and Wek “Eating te Meek" PE Dig Wash Oas, the 987 SHARD Dring Conkene, New ‘Grea, Loans, USA, ye 1-1, 1907 ser SPEAADC 16114 peste 3k 10. Orban, Demin MS, Jen BM ad Mayes New Uitasonie Clip or sD Opens” pape SPEIAD 21947, presented tthe 19 SPEADC Dri Careence, Artery, The ethane, March Ica, 195 October 1991 Example of ‘Anadil's SPIN sticking pipe inai- ‘cater analysis. From MWD down- hole weight-on-bit ‘and torque mec: ‘surements, surface hook load act torque, and a ‘model of the well trajectory, tho ‘analysis outputs both drag and rotary tiction fac: foxs dight track) As drilling pro- ‘coeds, preventive ‘action Such as ‘short wiper and reaming tripe are ‘made each time tho irletion factors increase beyond what experience dooms sao lias. ‘The MWD gamma ay data permit ‘cotroation of stick- ing events with ‘permeable sands Cow counts) and ‘Shales (high counts) 2BStenaot! “Tool diameter Standott Frequency Ld “A recon! MWD innovation—ultrasonic sensors that measure stemdofl othe ore- hole. An estimate of borenole diamotor i= obtained by summing two standoff moa ‘surements mace on opposite sides of the ‘MVD collar and adding the diameter of the MWD tool. Transmitted to the surface, the data are accumulated overtime and displayed in a distribution plot. This ‘shows the range of borehole diameters ‘encountored as the dilstring turns in the Iorehele. Smail and large diameters are [picked trom the lower and upper end of {ho disteibution and presented in log form versus depth. “Comparisons with wireline caliper logs run days later can reveal deteriorating ‘boreiele conditions that may presage the pipe geiting stuck. Botwroon X050 and X150 it, c shale section is collapsing. The ‘wireline calipers show that the hole has ‘enlarged since tho MVD calipers passed the zone and indicated the beginnings of ‘washout. In¢ permeable zone between 1X170 and X190 f, tho MWD data indicates Ihe hole tobe on gauge, while the wire- line logs show undergeuge hole. Tats dit. {erence is caused by mudeake development 24 Targe alameter | Borehole | 1 diameter | =| Mwo Wireline 2] uttasonic Mechanical 8] calipers calipers 2 in ei hoo} ] developed by Anadtill. As before, inputs to the friction calculation include hook toad and well trajectory, but the program also takes ino account mud buoyancy and BHA dimensions. Thus, the resulting fiction fac tor is independent of drillstring configura- tion and mud, an important consideration if these are being adjusted to cope with a sticking situation. The friction factor is is played on a rig floor monitor, providing an immediate indication ofthe onset of sticking." Ailer MWD technology, the computerized collection of dling data from surface sen: sors promises the greatest impact on fighting stuck pipe. The MDS information and alarm system being installed on Seco Forex rigs displays all pertinent data on ruggedized Fig-floor video screens for the dill and on worksiations to other key decision makers ‘on the rig, overcoming communications barriers. The system also addresses possible lack of communication during crew changes, because it allows amiving crews to easily review the previous days’ data, But ‘most important, the system alerts the driller to abnormal data wends and issues smart alarms so action can be zaken to avert crises. An example in fighting stuck pipe is the hook load profile.!2 During previous trips and drilling, hook load is stored and averaged in the system ach time the drillstring hangs free or is moving within defined speed! limits. Dis- played versus depth or time, this average hook load profile represents the expected trend when there is no sticking, This is com pared with the hook load trend recorded in real time during a trip. Any overpull is Immediately apparent on the video screen, and in case a hook load increase goes tunnoticed, a smart alarm automatically alerts all personnel on the rig (next page, above). Alarms raised on previous trips can bbe permanently displayed to warn of possi ble trouble to come. Further analysis can be performed on a workstation by comparing ‘overpul sections to lithology, ROP and dog. leg severity The full power of computerized informa tion systems has yet to be realized, though, bout a thorough analysis of data acquired during a recent sticking incident on a Sedco Forex cig drilling for TOTAL in offshore Bor: neo, Indonesia gives a flavor of the sophis cation to be expected inthe near fture ‘Around 2 PM. on Apail 29, 1991, drilling had reached 2066 m 16778 it) ina fairly Oilfield ReviewUnstable claystone formation. Because pene tration rate had dropped, it was decided to make a trp to replace the bit. Afr pulling ‘one stand, the pipe stuck on slips despite an ‘overpl of 75 klb (bottom, right). Using the rigs topdrive, it was possible to ream out for half an hour although circulation was blocked. But then the torque required 10 ream exceeded the topdrive capability, and the drilsving became truly stuck. A few days were spent backing off and connecting high shock jars to the stuck BHA. A¥ler consider able jrcing, rotation and circulation were re tesablished and the string became free. Detailed analysis of drilling data for the hour petiod preceding this event gives some indication that friction is increasing in this case probably caused by unstable for ‘mation that is falling into the hole and pack- ing off the BHA (next page). One indication is the hook load averpull experienced while drilling each time a new stand of pipe is connected to the dillsring—connections are indicated by hook load going to zer0 as the string is hung inthe slips, instantaneous overpull averaged 34 kb and rose to 54 Kb at 13.40 hours. Anather indication of increasing fiction isthe sporadic spikes on the torque log, which occur more frequently near the end ofthe log. ‘Monitoring standpipe pressure and torque during the last 3.5 hours and comparing it ‘with an earlier one-hour period yields fur ther indications that borehole conditions are deteriorating, During the earlier period, bath torque and standpipe pressure are rea sonably uniform. But in the later period, there are several pressure pales of up to 50 psi, These are thought to indicate temporary blocking ofthe annulus by sloughing forms: tion. Sometimes the pulses correlate with spikes on the torque log indicating simulta neous grabbing of the BHA. Expanded logs for the confirm that sticking is a likely problem Before attempting to pull out ofthe hole, the driller pulled up the string over the length o a stand and then ran the sting back in the hole. The difference in hook load between these two events gives a belter estimate of 7 ile Maya Zand Choo OX soe ose At ip Monn paper SPEAADE 2195, preset a the 1991 SPANO Deli Coneence ts th, the Sera arch 1,181 des Taam Ding a unay 1990) 15, 12, Boni ges Oi Revie? October 1991 AL 18 S005 on cae rae ee mn ad Display available in realtime on video screens of the MDS diling information ama alarm systor trom Sedco Forex, showing hook load (lett irack) cid other parameters, ‘versus depih as pipe is pulled. The yellow hook load curve built during previous Gilling and tips is the norm fo which the current, green hook load profive is compared. ‘The display shows the status when the bit has boen pulled fo 718.9 tho cuzzent ‘hook load fs closely following the norm. But eclior, whon the bit was between 3000 ‘and 4000 ft, there were sovoral excessive overpulls. These would have raised alarms {or the diller. To warn of problems, alarms irom previous trips aze superimposed on the screen—s0e the claem at 3700 ft. ‘Other information displayed includes trip tank volume (PBT V), instantaneous and cumulative mud volume balance (M VB/GUM), used for kick detection, and string ‘spoed (ST SPEED) VY seins sen "| / Gawd ; Ww “Analysis of a recent sticking event while pulling cut of hhole—Sedco Forex was drilling for TOTAL in olishore Born00, Indonesia. Displays of hook load and elevater postion versus lime show two stands successfully Pulled, and then the drill sting becoming stuck, with successive overpulls of over 90 kb failing to budge it. The rapid fall In hook Joad during the noxt fow nuinutos results from the formation bearing progressively ‘more of the weight of the BHA as it becomes trapped. 25Hook oad Hook oad 00 : 9 . - wh te 2 2 Pulingup_ ning m0 | a Gisting back nhole ‘to ! ‘0 ore oa ~ ~” Tage cue 22 « “ee Largo of: ' 2 s bottom torque 2 2 oT : ° | voto G700 0600 oO 1000 aD TaD ton oo wo OT to OM Se We = Tine he Wines alter 14001 standpipe presre standpipe presire 2700 2700 Pgosuo pos 2729) mrertcnt yantnrwrnmnveninntrrel goo 200 J exe Torque Tore © “po Spikes a * a \Driling data in the hours end minutes prior to the sticking event displayed in the previous figure. A detailed display forthe last ten minutes indleates a 23-kb overpill caused by drag friction and large torsional oscillations caused by the BHA alternately sticking {and slipping. Displays of standpipe pressure and torque for the perlod botween 12.6 and 14 hours shows pressure pulses causod by temporary blocking of the annulus and torque spikes caused by the dillsiing sticking, These events sometimes correlate. Compare with an earlier period, between 7 and 8 hours, when response is normal and no sticking is occurring. foverpull than the increase in hook load experienced at connections. The difference js about 23 kib, sil a considerable amount And indicative of large frietion losses some- ‘where in the borehole. Two others signs of increasing friction come from the expanded torque log, One isthe large oscillations near the beginning of the ten-minute period, caused by the BHA alternately sticking and slipping. The other is the large torque at the tend ofthe ten-minute period during which the deilstring was off bottom 26 This analysis would have been impossible ‘without the data gathering ability of a com- puterized information system. The challenge for researchers and engineers is to codify stuckpipe indications and manufacture ‘more reliable alarms for the diller This is an exciting time for such developments, Rig Sensors are improving, rig-based informa. tion systems are becoming accepted and stuck pipe research continues, As these forces converge in the next few years, stuck pipe incidents may become rarer, ensuring more costeffective drilling, HE Oilfield ReviewTechniques for Breaking Free Prevention of 1uck pipe is more economic than even the best of freeing procedures. But when prevention fails, the operator must move swiftly to select the most effective treatment. Here are some basics of unsticking pipe—from pills to jarring to fishing—that may help speed that decision. Denny Adelung, Warren Askew Jaime Bernardin AT. (Buck) Campbell, Sugar Land, Texas, USA Mike Chaitin Oryx Energy Co. Houston, Texas, USA Guy Congras Port-Gentl, Gabon Rodney Hensley British Petroleum Exploration Co. Houston, Texas, USA Bill Kirton British Petroleum Exploration Co. “Anchorage, Alaska, USA Randy Reese Don Sparling Onyx Energy Co, Tulsa, Oklahoma, USA 7 Freeing stuck pipe involves finding the right ‘combination of gentle persuasion and brute force. Gentle persuasion involves changing downhole conditions to reduce sticking so the pipe comes free on its own or is more responsive to force. And force is used when the hole’s grip on the pipe cannot be coaxed loose and must be overcome by a ipreater force. Force is delivered by pulling on the pipe, slacking off or applying and holding torque. Stuck pipe can also be banged free with a Jar—an impact device installed in the drill String that delivers a blow up, down or both to free the pipe (see “Jars, Jarting and Jar Placement,” page 52). Repeatedly opening, and closing the head valve in the mud ump several times causes pressure surges in the mud flow that may also vibrate and shake the pipe free. ‘The tools of persuasion are spotting fluids hole conditioning and changes in hydro- sialic pressure. A spotting lui is a volume, (or “pill” of fluid containing lubricants and surfactants pumped downhole and placed in the annulus opposite the interval of pipe held by the wellbore, the stuck point.’ Diesel ol products prove successful half the time and are used most commonly, but environmental agencies have recently encouraged use of less toxic mineral ols? bisa, PRT Preaure auton a Rl Tne) ie imukot nid Fer pn preparation of his article harks on ‘uli Bish Peocu Explraton, Hous eas (GSA ane eausegneu and Pee Puen, Sock osona’ HrAqueae, Pa, Pace in Bele, Hee ures, Chale nl, obs Kes, oval aston, Mare Lesage, Sah Mal, ry Pedi, ab Ferg and bert Fur, Anu Supa Lad, Toes, USA Roger Cancer C. Canaitats, Hume, “ous USA: Meahan Kala Wan Kal eating, Sar an, os, USA Some major oil companies have abandoned diesel altogether. Regardless of which base is used, spotting fluids contain additives that tenable them to act as both a lubricant to hp slide pipe around bends and through Light spots, and as a surfactant to weaken ‘cohesion between the mudcake and pipe ‘Occasionally, in carbonate formations a pill (15 to 25% by volume mud acid is used to parially dissolve hole bridges, cuttings o sloughed formation. Hole conditioning once stuck involves Increasing mud flow rate or changing mud properties such as density solids and pH) to tenhance solids removal, such as a cuttings bed in a high-angle well ora plugged val. Reduction in hydrostatic pressure is used mainly to free diferentally stuck pipe. This can be achieved by lowering mud weight or the level of mud in the annulus if well conditions permit—no risk of gas in- flow and good hole sibility. An altemative js to "U-tube" the drillpipe and annulus, leading to drainage of the annulus and, more elaborately, backing off and installing a drillstem test (DST) tool (see "Freeing Stuck Pipe: An Overview of Two Tech: iques,” next page). An increase in hydro- static pressure, by caising mud weight, is used to treat sticking in unconsolidated, mobile or geopressured formations. cominasdon page 30) 1. Fortcniqus of pl placement: Grogan GE: "How tofee Suc Onl Pipe," Ol & Ga ur ro, Caan 1966) 18-8 2 Ayes RCH ORely and Hey LR: “Orbe ‘Oper Commie Cal Sete Spin ‘Srey pape SFUMAI 1968, peed the 1969 SPEADC Ding Coreen, New Os Bulag, USA ebay 28-Atrch 3, 1909. Shir Raed Domargie i “Opeations Deion Mating fe Stuck Pipe Ince inthe Gul of Messen ‘Nik teenies Arorvch per SPUIADC 21998, eet te 191 SEO Bling Coe ented, The Retherad, Mac 14 1991Zc Ditrenal sticking «Sticking ocours when ditsting is ‘stationary + Dlisving cannot be rotated or moved up or down orf sting ‘can be moved, high ovepul at ‘connections and wile tipping, and high torque night after a connection. + Generaly no signiteant pressure vations in creulation + Porous and impermeable zones, present in tho openhole interval May be a depleted zone 1. Stop or reduce circulation to lower ‘effective mod weight at stuck pont. Pull with maximum allowable force. ‘Start faring at once at maximum force. Jar upward for frst hour then both ays. Ty olan, Make eon- ‘gation measuromonts wile varying pipe tension to estinato location of stuck point. 2.As soon as possible, add surfac- tant sg whl working drilsting Be sure to give the adctve time to Work bofoce washing out the sug 8 if previous attempts fal holes sta- ‘bls and thee sno rk of gas Jnlow (back off ac doop as poe- se, tuna with a DST tool and Screw ito fish. Open tester fr ‘init; donot tow we. Tis ‘aleve coronal pressure on fr ‘mation under no-fow condition.” 1. Start working the pipe immediately Ifthe Bits of bottom, slump the pipe with night nan torque nln {tif the bit sen bottom, pul to the ‘maximum immecatly ad. ‘Apply iht-hand torque to try to get ‘movement to the stuck point 2. I rst attempts fal work the pipe ‘n both erections untl atematve ‘action can be taken. 3. Consider ll aspects of well conta ‘oforelotering nyorostaticpres- ‘sure In choosing how to lower ‘reseuro, romomber the dogroo of ‘coi are speed of the pressure rodustion, Pressure reduction f most effective the drilsting isn ‘compression. 4. Locate the stuck pont to ensure ‘the pli of sutfictnt volume to ‘cover the stuck point. Fluid should (02 ppg heavier than the mud to ‘ensure that t stayin lace. Put the pe inte compression, stacking off 0,000 b below the weight of ho pips an apply a i tu of ight hare torqua for ovary 300 m (1000 fof pipe above the suspected ‘stuck pot Every 5 minutes release torque and pick up the 10,000 b.Laave a reser volume of il inside the cipipe and every Salminutos slowly pump haf @ber- Telto move moc pil into the hole. ‘Worth pipe up and down a reg- iar intervals. low 12 hours of soaking time; 40 hous ison considered excessive 5. The U-tube mathod is an atema- tive requiring care wal contol? Mechnical sticking Koy seat ‘Ista cola or stabiizor caught at ‘a specie depth only when tipping ‘out + Froe passage at suspected key ‘eat depth when running in + Stuck point conesponds to doaleg + Tight hole will occur at ho same ‘depth on tps. Dificul to ently special, ‘Comment caused by lack of precautions whe ning intel, * Distring too rg corned wit previous sting. ‘New bit larger diameter than worn be + cast sol bit run ater a iicone bt ‘Never pal strongly: ry moving down ‘wth torque. Jar down. Spot 2 hgh IMreating sia, if unsuccessul back ‘ff as close to stuck point as possible. unin wih aring sting of ouside ‘larotoremal enough to enter the key seat on running i, 2 09 hours of ring is unsuccessful, spot ib. ‘ant and continue jing. If ferential Sticking occurs, back oft the pipe above Key seat. Hold fish wih spear tw prevent itfaling and washover Vibrate pee with mud pump while ‘moving in rection opposite that ‘when the pipe stuck. Usually ar ‘down, Set libricant and continue. Emphasize jaring use same proce dures a for fallen objec ‘Work the pipe down. When the pioe can be worked down andrtated, Dut cannot bs pulled past the key seat Stowly rotate the sting with mma tension. tis hoped this wil work the Colas and stabilars past tho Koy seat ‘Werk te sting an arn both directions Increase forces gradual Work sting upward and jar upward ‘Apply maxenum forces from the sta. Work string upward ane jar upward, ‘Apply maximum forces fom the sta.