E *... Society of Petroleum Engineere ladc/spe 39354 Trends extracted from 8 Gulf Coast blowouts during 19-199 Pal Skalle/NTNU, Trondheim-Norway; Augusto L. Podio/University of Texas at Austin, Austin, TX-USA Copyr~t 199S,lAOC/SPE DrllllrrgCwferenca This peper was prepared for presentation at Iho 19SSlAOO/SPEDrilling Conference held in Dallas,Texas H March 1S98. This paper MS wlected for presantatl~ by an laoc/spe Program Conunittee following ravlew of informalhm Mtalnti In an abstract autitfad by tha author(s). Wtents of the paper, aa prasentad, have not bean retibward by the Intamatiaal Aaaoclallon of Orilling Contractors or the SooIaty of Petroleum Engineers nd ra sub@ to oorrnotiur by the author(s). The material, aa presented, ha not n-aaarily rafld anypositionofthe l= or SPE, their officers, or mambara. Papera prsaenled t the IANSPE -Iings are subjeef to publi~tiw review by Editorial Cmltteas of the IAOC and SPE. Electronic rapreduotim, distributi~, or storage of any part of thla papar for an?arcial purposes withouf tha writt~ wsent of the Sooiety of Patroleum Engln-ra is prehlbited. Permlaa[onto r~oduee in pdnt is re8triofw 1 n abatrscf of not mere than 33 ~; illustrations may not ba copied. The abatraof must wntain maplououa aolmowedgrnent ef *are and by *em tho paper was presented Write Librarian, SPE, P.O. Box 8338W, Rtiardaon, TX 7S33-3e3, U.S.A, fu 1-972-952-W35 Abstract The cost of blowouts and the loss of life incurred from blowouts warrant the study of past occurrences in order to reduce number of future blowouts. Even though we continuously learn more about how to handle an unstable well, it seems that the problems of detection, handling kicks and loosing control do not change much with time in the real world - out on the rig, The small changes and improvements that do occur are first of all seen thorough statistical data. A new data base resently compiled will soon contain extensive data on more than 8 blowouts from the Gulf Coast area and adjoining states duting the period 19 through 199. At the moment the compiling work has not been completed. However, at the 1998 [ADC/SPE Drilling Conference an updated version of this paper will be available. Five agencies contributed blowout report information. The State Oil and Gas Boards/Conservation Office of Alabama, Mississippi and Louisiana, The Railroad Commission of Texas and the Mineral Management Service (MMS) of the Dept, of Interior. At the moment the Texas blowouts makeup approximately 5Vof the blowouts in the data base but range below the average blowout frequency of.15 blowouts pr. 1 well-ft. drilled. The three major operations in progress when blowouts occurred were exploratory drilling, workover and development drilling. Most frequently blowouts occurred during the activity of actual drilling, tripping out and circulating/killing. The analysis indicated also that the major causes were related to swabbing, drilling into a high pressure zone or formation break down while the secondary barrier after a kick has occurred which most often failed were either failure to close P, failure to stab string valve or P failed after closure, Introduction Control of the formation pressure is an important concerr both in the planning phase and during execution of the drilling operations. Improper procedures with respect to detection and handling of kicks may result in loss of control. The cost of blowouts and the loss of lives incurred from blowouts warrent the study of past occurrences in order tc reduce the possibilities of titure blowouts. We are hoping to see that operating experience in general and experience through required blowout schools especially is slowly paying off. Technology changes may however override these effects in both directions. Examples 1 technology changes are diverterless drilling combined with drilling in deeper water and pushing the surface casing deepel and deeper/fewer casing strings, thus causing a changing risk and not necessarily accompanied by a corresponding reassessment of the kick detection/killing procedures. Since the shallow gas hazard and consequences are above normal, large improvements have been noted the last decade or twc with respect to shallow seismic and kick detection systems for floaters. Method of Analysis On basis of data from five data sources State Oil and Gas Board of Alabama Louisiana OffIce of Conservation Mississippi State Oil and Gas Board Texas Railroad Commission (RRC) Minerals Management Service (MMS) (Outer Continental Shelf) 539
2 PAL SKALLE, AUGUSTO L. PODIO SPE 39354 we have created a new data base in Excel spread sheet for the period of 19 to date, and plan to update it every 5th year in order to watch over new trends within blowouts. The database was initiated by Hughes, Podio and Sepehrnooril. It was not until 1973 that mandatory reports were instituted in all the US states. Before that time no systematic method of record keeping existed, From the investigated records, both before and after 1973, we have seen that sometimes kicks are reported together with blowouts. Kicks which have simply been circulated out in ordin~ manners have been excluded from the data. Those kicks that have caused extraordinary problems during shut in or killing, like loss of circulation, stuck pipe/underground communication etc. are included in the RRC data base, and it has been somewhat difficult to determine if it was a blowout or not. When in doubt we have included such occurances and denoted them with very short blowout duration. In addition to drilling operations (sulphur drilling is not included), blowouts that occurred during completion, workover and production are included. Therefore our data show higher frequent ~ than drilling operation blowouts published by Daneberger. Wells drilled before 19 were also excluded from our data base due to lack of accuracy and consistency, Reports on wells drilled after 19 show a gradual increase in both consistency and detail level, although some of the parameters had to be interpreted, even after 1973. The data base contains too much data for one paper; In order to avoid a voluminous paper we decided to publish statistics dealing with what happened after the blowout (blowing fluid type, mode of control, duration, consequences etc) in a later paper, Drilling Activity vs Blowouts To date (1997) some 8 blowouts have been implemented in the data base, split on areas as shown in Table 1. Table 1. Overall activiy and no. of blowouts between 19 and 199 Area No of No of wells Footage Blowoutspr. blowouts drilled drilled 1well 1oft Alabama 9 ** ** ** I ** ~.. * Estimated value before all data have been com~iled ** Still missing data Activity distributed into different categories are seen in Figures 1 through 5. Figure 1 and 2 show that the drilling activity had a peak in US early in the 8 s, corresponding to a high level of blowouts. Table 2 indicates that the overall blowout frequency is rather stable, and independent on the activity level. Table 2. Blowout frequency in Texas (incomplete data) Ycnr 7 71 72 73 74 7S 7 77 78 79 $ Blowouts,1,3,M,7,11,M,3,5 9 8 i pr. 1S wells Year 81 82 83 84 85 8!37 88 89 9 5 Blowouts,9,7,5.8,7,M.5,1,8 8 1 Dr,loo Blowout depth Number of blowouts were compared to actual depth the kick ocurred at and the last casing set in Figure 3 and 4. It is obvious that most kicks occur at shallower depth since most wells are drilled here, It is almost even as obvious thal blowout frequency is much higher in deeper wells since the &llling length/exposure time is higher and formation pressure is higher. When compared by drilled length within the depth categories this evens out, but as expected, the frequency is highest in the deepest depth categories due to higher pore pressure gradients and difficulties in handling highl compressed gas. In addition, as stated by Wylie and Visram K, increased exposure time, longer open hole sections, more tripping time and increased risk of lost circulation problem: increase the blowout probability. Operation in progress Operation in progress have been devided into two groups, first the operation have been subdivided into phase of operation (drilling, completion, production, workover and wire line), secondly the major phases of operation have been further split into operation/activi~ in which a blowout has occurred. The analysis is presented in Figure 5 and Tables 3, 4 and 5. Most risky operational phase is exploratory drilling where the blowout frequency is 3 times higher than for development drilling3 4. During the drilling phase it is mainly in the drilling operation the blowouts occur, more precisely activities like; drilling into unknown geology, tripping /axial movement of the string and being out of the hole. The high percentage of blowouts during drilling and tripping indicates that operatorslcontractors are not taking enough precautions to compensate for the decreased overbalance margin. 3 The industry must walk a fine line ir choosing a fluid that will optimize both kick control and drilling rate efficiency. Blowout causes Blowout are failure to control a kick and are caused either by equipment failure or human error. A blowout occttt approximately for every 11 kicks that occur3, both fo~ develop ment/explora~o~ wells. ~n--tabiei throug~ 9 the causes of blowouts have been systematised, and we will let the 54
SPE 39354 TRENDS EXTRACTED FROM 8 GULF COAST WOUTS DURING 19-199 ~. data speak for themselves. Reducing number of blowouts In spite of improved regulation and inspection of P equipmentirigs and improved blowout prevention training for drillers and wellsite supervisors no improvement in blowout frequency in Texas was seen during the years 1978-1991, in Table 2, This can be explained by the fact that the prevailing contracts were of the footage or turnkey type. Improved rig equipment and crew ability to detect and control kicks were lost in contractors effort to maximize revenue through increasing the drilling rate (minimum overbalance) and/or decrease the non-productive rig time (ignore or missing drilling breaks, tripping too fast, not stopping to circulate bottoms up to remove trip gas or to check for any minor influxes or more frequently fill the hole). Wylie & Visram3 suggested to combat such trends through Eliminate footage contracts in favor of day work contracts. Alternatively operators should take a more proactive role in specifying minimum mud density and offer a bonus program for minimizing kicks. Take the same approach to kicks as for rig/equipment inspections. Each kick should be submitted to the Conservation Board and made part of the public record for review, analysis and continuos improvements. Conclusions A reasonable extensive data base containing blowout data from Texas, OCS and US Gulf of Mexico States are about to form, containing data from 19 to date. The data quality is improving, especially after 1973 when mandato~ blowout reports were instituted in all US States. Like before blowouts are occurring for two main reasons; 1) Drilling into a formation we thought we knew, but the fomlation pressure was higher than expected, combined with the effect of low mud weight amplified by means of gas cut or water cut mud and axial movement of the string causing swabbing. 2) Too high mud weight/high ECD during circulation or downward movement of the string leading to fractures, loss of circulation and kic~lowout. Although the.-. first hurrier is lost in such nlanners the situation co!ld haye been controlled if the P had been closed, would have worked or was installed. It sounds simple, but the real world is more complex than these simple rules. References /1/ Hughes, V.M.P., Podio, A.L. and Sepehmoori, K.: A Computer Assisted Analysis of Trends among Gulf Coast Blowouts, In situ, 14(2) (199) 21-228. /2/ Daneberger, E.P.: Outer Continental Shelf Drilling Blowouts 1971-1991, OTC paper 7248, Proc.: 25th Annu. OTC, Houston (May 3-, 1993) 415-425. /3/ Wylie, W.W. and Visram, A. S.: Drilling Kick Statistics, IADC/SPE pa~er 1994, Proc. at the IAI)C/SPE- Drilling Conf., Houston (F~b. 27-Mar.2, 199) 77-8. /4/ Hohmd, P.: Offshore blowouts, causes and trends, PhD-thesis, 92, NTNU, Trondhehn (Feb 199). Table 3. No of blowouts (J vs o~eration/activi@ in progress during the drilling phase ~x + OCS, 19-9) Operation Drilling 29 Circulation 42 Technical problem 1 ActiviQ Actual drilling 95 Tripping outicnxfwiper 82 Out of hole 13 Tripping in 5 Coring 3 Install P 1 Nipple down P 1 Circulatingikilling 2 Test P 3 Wait on order 2 Fishing Stuck pipe 5 Killing 2 Snubbing out 1 The problem of detection, handling kicks, loosing control does not seem to changemuch Over!he years. The improved technology and improvements in procedural abilities is lost in the strive to maximize efficiency and revenue, thereby overlooking vital pieces from the fine set of small indicators, especially if the casing program is challenging the geological nature. The answer to this complex problem may be found in enhanced motivation and knowledge. Who knows. 541
>, 4 PAL SKALLE, AUGUSTO L. PODIO SPE 39354 Table 4. No of blowouts () vs operatiotiactivi~ in ~rogress during the completion phase (T.x+ OCS 19-199) Operation Installing equipment 25 Well testing incl. 1 preparations Circulation 1 Activity Nipple down P 5 Woc 5 Casing running 3 Cementing casing 2 Fishing 1 Stuck pipe 1 LOT 1 Set well plugs 1 Woc 5 Cementing casing 2 Tripping in 1 Tripping out 1 Squeeze cementing 1 Killing 2 Perforating 1 Cleaning well 1 Gas Iifiing 1 Table 5. No of blowouts () vs operatiodactivity in Drogress during th~ work over phase {Tx +- OCS; 19-19b) Operation Pulling well equipment 37 Installing equipment 17 Abandon well 1 Activity Pull tubing 15 Stuck pipe 4 Pull/drill out well plugs 3 Pull WL 2 Logging 2 Perforating 1 Cleaning well 1 Snubbing out 1 Run tubing 5 Install P 3 Run WL 2 Nipple down P 1 Set well plugs 1 Acidizing 1 Pull tubing 8 Set well plugs 4 Killing 2 Nipple down x-mas tree 1 Pull/drill out well plugs 1 Table. Most frequent prima~ and secondary barriers that failed in allphases (Louisiana + Tx + OCS; 19-199) Primary barrier 1 Secondary barrier Swabbing 158 I Failed to close P 78 Too low mud weight 5 Rmrrsnot seated 14 Drilling bretiunexp, high pressure 45 Unloaded too quickly 13 Formation breakdotiost circd, 43 DCKeIlyiTJ/WL in P 5 Wellhead failure Trapped/exparrdinggas Gas cut mud x-mas tree failure While cement setting Urdmom why Poor cement Tubing leak Impro~r fill up Tubing burst Tubing plug faihue Packer leakage Amndsr losses Uncertain reservoir depth/pressure 4 4 33 23 2 19 1 15 13 1 9 P failed after closure P not in place Fracture at casing shoe Failed to stab valve/kelly~iw Casing leakage Diverter - no problem String safety valve failed Diverter failed after closure Form, break downilost cirmd, String failure Casing valve failed WeOheadseal failed Failed to operate diverter x-mas tree-failed 43 38 34 23 21 19 17 15 13 II 1 7 7 Table 7. Distribution PA) of most frequent operation phase failures (Tx + OCS; 19-9).. ṃ )Uts primary barrier Swabbing To low mud weight Drilling break Formation break down Wellhead failure Trapped/expaudirrggas Gas cut mud x-mas tree failure While cement sets Secondary barrier Failure to close P P failed after closure P not in place Fracture at casing shoe Failed to stab string valve Casing leakage Tx E 2 28 32 21 34 2 15 9 z 45 2 2 28 14 Ocs T 12 14 4 1 7 12 9 3 4 1 & 2544223 3312184 2312 424258 12834518 182813353 4231323 7129 11795 34 44 4 24 4 342713 ;13 1813343 155848 2 24 3 47 421743\4 542
SPE 39354 TRENDS EXTRACTED FROM 8 GULF COAST WOUTS DURING 19-199 ~ Table 8. No. of blowouts () vs primary/secondary barriers dz[ring the ~rillingphase (T.x+ OCS; 19-199) - Prim@ cause Swabbing 14 Unexpect well press/drlg break 41 rrappetiexpanding gas 39 Fomration breck down 3 Sas cut mud 27 While cement setting 19 Unknownwhy 1 Poor cement 1 [mpropertill np 7 Uncertainresemoir depth 5 4nnular losses 5 Water cut mud 3 Drilling into rreighbourwell 3 Most frequent secondmy cauaea Failed to close P 52 Rams not seated properly (4) Unloaded too quickly (1) DC, TJ etc. inside P (3) Failed to stab atring valve 19 P failed afier closure 11 Fracture at casing shoe 11 String safety valve failed 8 P not in place 5 Diverted no-problens 4 String failure 4 Annular valve failed 3 P failed after closure 9 Failed to close P 4 Diverted - no problem 4 Fracture at casing shoe 3 Casing leakage 3 Formation bresk down 3 P failed afier closure 1 Fractnre at casirra shoe 5 String safety val;e failed 3 P failed afier closure 7 Failed to close P Fracture at casing shoe 5 P failed after closure Failed to close P P not in place 3 P not in crlace 11 P failed ~fter closure 5 Diverted - no problem 4 Diverted - no problem 3 Failed to stab string valve 2 Fracture at casing shoe 2 String failure 2 Diverter failed 2 Table 9, No. of blowouts () vs. primary/secondary barriers during cornp[etion and workover-phase (Tx + OCS; 19-199). Prima~cmme Wellheadfailure 27 X-mastree failure 19 rubing to anmdus leakage 13 rubirlg pl~tgfailure 7 rubing to cnmdus burst 7 Casing plug failure 5 WL lubricator faihlre 3 SCSSV/Storrrlckoke failure 2 Snubb,eqtlipment failure 2 Casing collapse 2 Most frequent secondcry cause Casing valve failed 2 Wellbead seal failed 3 Casing leakage 5 543
5 PAL SKALLE, AUGUSTO L. PODlO SPE 39354 8 v o ~ ; 4 I-R---USA /\ 195 197 1975 198 1985 199 1995 Year -.... ~igure I We[[s dri[[ed in Texas, offshore (Texas and Louisiana) and US by nuntber CIfwells.... -.... 25 o 2- - w % o g15- - x E x OTX x= ~lo- - oxxxoo X. xx 2 x Oxxxo x ox 5- - OXXX5 Ax?:., x x, ~~ --, - 19 197 198 199 2 Year... uimre 2 NO of. blowouts in Texas and OCS bv,. vear.
3PE 3934 TRENDS EXTRACTED FROM 8 GULF COAST WOUTS DURING 19-199 7 45 4 = - 2 %. 15 2 1 5 o 5 1 15 2 25 Depth (1 ft) ~i~r= 3 NunzbCrof b[~wouts in Texas and OCS by well depth during 19-199..-... 12 1 TX rl HOGS ḣ.-. -: 2 n 1 + A Casing size (in)....,..,... ~iwre 4 N~nlber of b[owouts in Texas and OCS b-vcming size categories during 19-199.
..,,,.. 3 PAL SKALLE, AUGUSTO L. PODlO SPE 39354 14 I 12 1 8 4 J 2 -.-., -.-.- -...,.----- -----.. L ~igure 5 Blowot~ts vs phase in progress flti OCS). i