Applying Technical Limit Methodology for Step Change in Understanding and Performance

D.F. Bond, SPE, P.W. Scott, SPE, P.E. Page, SPE, and T.M. Windham,* Woodside Offshore Petroleum Pty. Ltd.
Summary

This paper presents an alternative planning approach to the drilling and completion process, technical limit, which has resulted in a step change in Woodsides performance. Three new wells and six subsea completions were finished 20% under budget with this tool and with a simple philosophy characterized by the following questions. What is current performance? What is possible? What is needed to get there? The target was to drill a directional well in 20 days when the previous best time was 42 days. A target of 12 days was set on subsea completions, although a conventional approach had previously been 20 days. The methodology was to ask what would be possible if everything went perfectly on every operation making up the well time. This is not the usual trouble free time but a well time built up of individual components, with each component representing its theoretical best performance. Details of how the approach was used to plan, and operational data that confirm that the technical limit can be approached are presented. As a result, the well construction performance delivered step change improvement when managed against the technical limit.
Introduction

to every component of well construction resulted in aggregate drilling and completion time, which was called the technical limit. Technical limit was a term used to describe a level of performance defined as the best possible for a given set of design parameters. Such performance can be approached but requires a perfect set of conditions, tools, and people. A close analogy of the technical limit is a world record in athletics.
The Philosophy of Targeting Technical Limit

The decision to target technical limit was a profoundly significant one. As with a decision to pursue a world record in athletics, pursuit of technical limit required extraordinary effort and commitment and would probably not be achieved. This is contrary to the commonly held belief that targets must be achievable. Implications included a need for highly competent people, team work, and effective leadership. In meeting the challenge, many deficiencies were exposed (and dealt with) that had not been exposed in the past, thereby inviting criticism and invoking the blame culture, not uncommon in the oil industry. A key success factor was to stay committed to the technical limit strategy, not to image.
Determining the Technical Limit

Drilling performance offshore on the North West Shelf of Australia from 1968 to 1992 was erratic. A simple plot of time vs. total depth (Fig. 1) showed unacceptable scatter and a high average drilling time, particularly when benchmarked against published data (Noerager et al.1). The conclusion was the well construction process was not in control. The company developed a plan to remedy this when faced with an upcoming development project (Wanaea and Cossack subsea development). An aggressive target setting and planning methodology was developed on the basis of question, What is possible? rather than the question, What can be improved? to get the desired performance. The approach was greatly influenced by achievements in other parts of the world. Some new drilling operations standards were set in the North Sea during the late 1970s and early 1980s, as documented by Shute et al.2. A major factor claimed in the success of this work came from time analysis, which rigorously pursued the identification and removal of drilling problems. Work by Huber et al.3 in the North Sea took a similar approach, which also produced excellent results. The high level objective for the Wanaea and Cossack projects was the requirement for highly productive wells and low construction cost. Work carried out to maximize well productivity had the highest priority because of the potential pay back and has been documented by Walters.4 The focus of this paper was the reduction of well costs. With time as the predominant cost driver for well cost (70% of the well cost was time sensitive), time reduction took priority over unit cost reduction. Application of the What is possible? question

Copyright 1998 Society of Petroleum Engineers This paper (SPE 51181) was revised for publication from paper SPE 35077, first presented at the 1996 IADC/SPE Drilling Conference held in New Orleans, 1215 March. Original manuscript received for review 12 July 1996. Revised manuscript received 30 April 1997. Paper peer approved 17 April 1998. *Now with Chevron Overseas Petroleum Inc.

The Theoretical Well. The first stage in the development of a technical limit well time was the construction of a theoretical well. The theoretical well assumed a flawless operation on the basis of current knowledge and design technology. It was made up of activities and durations that were derived from collective experience. Assumptions included, for example, no midsection trips required to change bit or bottomhole assembly (BHA), no reaming (stable hole), no waiting on equipment, no wiper trips,2,3 and no significant circulating time. This diverged from the assumptions made by Kadaster5 for a normal well. We concluded that to include such times would have included technical shortcomings in the plan. The goal was to highlight technical shortcomings as the focus for action and change, rather than accept, them. The well construction was broken into easily definable sections to calculate a theoretical well, such as drilling a 1712-in. hole, running and cementing 1338-in. casing, and drilling a 1214-in. hole. This produced about 9 to 16 sections depending on the design detail of the particular well. The sections were then broken into subactivities as shown in the following example. Section: 1712-in. Hole. Lay down 26-in. BHA; pick up 1712-in. BHA; run in hole; drill shoe; perform leak-off test; drill 1712-in. hole; circulate bottoms up; and, pull out of hole. The aim of breaking each section down was to define sufficient design detail to enable estimating of actual durations; too much detail became cumbersome to work with. Durations were then estimated for each subgroup. Drilling time assumed a 10 minute connection time3 and rate of penetration (ROP) from the best bit runs in the area. A well time produced by this method often resulted in disbelief of the short duration calculated (see Table 1). By developing the theoretical wells technical limit in a group session, it was possible to bring about a shift in paradigms about what was, and is possible. Removable Time. The difference between actual well time and theoretical well time was called removable time. It included con197

SPE Drilling & Completion, September 1998

Fig. 1Days to total depth (TD) for wells drilled on the North West Shelf between 1968 and 1992.

TABLE 1ACTUAL VS. THEORETICAL DURATIONS Theoretical Time (Days) 22.0 22.0 Actual Time (Days) 42.3 51.5

Offset Directional Wells Cossack 2 Cossack 3

ventional lost time and down time and also another component termed invisible lost time. Invisible lost time was the name given to the time taken to perform those activities included in a normal well but excluded in the theoretical well. It was called invisible because, before development of this method, it had not been recognized or reported as lost time. Fig. 2 shows the relationship between theoretical well, invisible lost time, and conventional lost time.5 The times being generated by the theoretical well were indicating that reductions of 40 to 60% were possible, so quantifying the amount of removable time was important. Identification of the removable time (invisible lost time and normal lost time3) was achieved manually by use of daily reports from offset wells and application of the assumptions made for the theoretical well. The exercise of extracting removable time analysis was very time consuming (up to 2 man months for the eight wells reviewed) and required a high level of drilling/completion knowl-

edge. The study provided quantification of the missing invisible lost time and a list of all the problems and technical limitations that prevented best performance on those wells. The result was significantly greater after subtracting removable time from actual time than the time predicted by the theoretical well. This indicated that either the theoretical well was overly optimistic or that further removable time was hidden in the reports. To deal with this discrepancy, another term was introduced, effective time (shown in Table 2). It was concluded that there must be a component of invisible lost time associated with such things as efficiency improvements in rig crew activities and drilling optimization (e.g., bit type and bit weight). It would require cross comparison of inefficiencies on different wells to identify the best. The data set of the eight immediate offset wells was checked for effective time for each section to investigate these inefficiencies further. Table 3 shows the effective time for each section of the wells. The highlighted times represented the best sections, and, when combined, they added up to 19 days. This compared very closely with, in fact a little lower than, the theoretical well times estimated, although design parameters had changed somewhat for the new wells. The conclusion was that the theoretical well times were valid as the starting point for the technical limit. The theoretical well time developed for the completions (as with drilling) also appeared too low. A similar exercise of removable time analysis was performed on a comparable North Sea subsea completion campaign. The results from the study confirmed the theoretical durations. An opportunity to reduce well times by up to 60% was revealed by objectively assessing the removable time with this method.
Use of the Technical Limit Method

Planning. The planning process used the technical limit concept to identify the problems that prevented the technical limit being

TABLE 2AN EXAMPLE OF EFFECTIVE TIME WITH COSSACK 2 AND 3 DURATIONS Theoretical Well 22.0 22.0 Actual Times 42.3 51.5 Removable Time 13.6 22.3 Effective Time 28.7 29.2

Offset Wells Fig. 2Diagrammatic representation of the relationship between the theoretical well, invisible lost time, technical limit, and conventional lost time. 198 Cossack 2 Cossack 3

SPE Drilling & Completion, September 1998

TABLE 3EFFECTIVE TIME IN HOURS FROM WANAEA AND COSSACK REMOVAL TIME STUDY Wanaea Section Run TGB Drill 36-in. hole Run 30-in. casing Drill 26-in. hole Run 20-in. casing Run Bop Drill 17 12-in. hole Run 13 38-in. casing Drill 12 14-in. hole Run 9 58-in. casing Drill 8 12-in. hole Log Run liner Clean out Total hours Total days No. 1 2.25 9.50 10.50 28.25 10.25* 13.25 133.25 32.50 270.25 44.00 130.25 No. 2 2.00 8.50 7.00 33.00 11.50 13.50 110.00 23.50 201.50 25.00 53.00 49.00 59.50 44.00 640.00 26.70 No. 3 3.00 11.00 6.00 23.00 13.50 13.00 118.00 36.25 119.50 27.50 72.00 71.00 25.00 37.50 576.25 24.00 No. 4 2.50 19.50 6.75 27.25 13.50 10.00 117.50 21.75 105.50 24.25 90.75 54.00 37.50 26.50 227.25 23.20 No. 5 1.50 7.50 4.60 27.75 13.50 16.25 114.75 25.00 106.50 44.75 71.25 36.50 41.25 No. 1 6.50 18.00 9.50 28.50 11.50 17.00 155.50 22.50 319.00 24.50 61.00 43.00 23.50 45.00 785.00 32.70 Cossack No. 2 2.50 9.00 10.25 30.50 13.50 9.00 172.75 22.75 122.75 42.25 98.50 22.75 42.50 69.50 668.50 27.90 No. 3 2.50 15.50 8.50 25.75 15.00 10.75 174.00 23.00 193.75 23.75 111.25 62.0 42.50 Technical Limit 1.50 7.50 4.60 23.00 10.25 9.00 110.00 21.76 106.60 23.76 53.00 36.50 23.50 26.60 466.26 19.00

684.5 28.5

511.00 21.30

715.25 29.80

* The highlighted cells identify previous best (Technical Limit) times.

reached and then planned their removal. The time taken to drill and complete wells would be the indication of the ability to understand and manage the variables. Approximately 9 months were spent planning the startup of operations. The planning process used the theoretical well and offset well analysis to develop an engineering workscope and determine the resources needed. Execution of the engineering workscope involved getting engineers and contractors to understand potential obstacles to achieving the technical limit and then manage them. Approximately 175 days were identified as removable time from a total of 435 days on the eight offset wells. The problems identified by analysis are shown in Fig. 3 as a Pareto chart. Significantly, the invisible lost time was categorized as bit/BHA, 47%; mud, 23%;

waiting on weather (WOW), 13%; and other, 15%. Clearly, if the bit/BHA and mud problems were addressed, then 72% of the removable time would be eliminated. Drilling and completing outside cyclone season would potentially remove another 13% because of WOW. The development of the theoretical well activities and times provided a baseline for further analysis of the sequences required. Program evaluation review technique charts were constructed, typically listing 300 tasks/well, and used to undertake critical path analysis. A number of activities were removed from the critical path by introducing new tools and/or techniques. The engineering planning work had a big impact on the specification of the drilling rig needed for the project. The cost of the higher rig specification was easily justified against the potential

Fig. 3Results of the removable time study performed on eight offset wells. Categorized by bit/BHA, mud, WOW, and other. SPE Drilling & Completion, September 1998 199

Fig. 4 Diagrammatic representation of the management model developed to support the technical limit approach.

efficiencies indicated by the theoretical well time. The result allowed fit for purpose rig selection and, ultimately, sole source negotiation of the selected rig rather than a low bid tendering process. A number of other factors were addressed during the planning stage and were influential in the overall success of the project. These included identification and management of risk; communication of the plan to all involved, to gain ownership and shared goals; and, preoffshore review of procedures, sequence of events, and equipment. Operations. Invisible lost time was made very visible in the operational phase of the project. Any activity time deviation from the technical limit schedule was reported on the daily report. If the deviation was negative (i.e., time reduced), the time expected on future wells was adjusted downward and a new technical limit was defined. If the deviation was positive (i.e., extra time taken), it was analyzed for cause. If the extra time taken was unavoidable (i.e., the

technical limit time had been underestimated), a new time was defined for the next well. If it was removable, solutions were developed that would prevent reoccurrence of the event. Similar to the approach taken by Kadaster,5 the use of a total quality management (TQM) approach was found to be very effective (see the TQM feedback loop in Fig. 4). The feedback system was very broad in its application and became very efficient with proper resourcing from either company or contractor. Initially, offshore personnel were uncomfortable reporting against the technical limit because every nonconformance, large or small, was exposed. It was important that management encouraged and supported the offshore team in pursuit of the ideal standards that had been set. A no-blame environment was essential and practiced with vigor. The measurement of operations against the technical limit schedule was extremely powerful. No better way was found to highlight and pursue maximum opportunity for improvement.

Fig. 5Days to total depth (TD) for wells drilled on the North West Shelf between 1968 and 1992. Results from Wanaea and Cossack have been included to show the improvement made. 200 SPE Drilling & Completion, September 1998

TABLE 4POST-WELL ANALYSIS FROM WELL 1 Starting Technical Limit 9.0 11.0 21.0 25.0 20.0 107.0 36.0 77.0 44.0 65.0 6.0 57.0 476.0 19.8 New Technical Limit 10.0 7.0 21.0 25.0 20.0 107.0 36.0 77.0 42.0 65.0 6.0 51.0 467.0 19.5

Section Drill 36-in. Run 30-in. Drill 26-in. Run 20-in. Run BOP Drill 17 12-in. Run 13 38-in. Drill 12 14-in. Run 9 58-in. Drill 6 12-in. TD logging 4 12-in. liner Total hours Total days

Actual Time 11.0 9.0 27.0 67.0 30.0 380.0 43.0 142.0 42.0 104.0 14.0 292.0 1158.0 48.2

Removable Time 0 2.0 0 42.0 0 249.0 2.0 48.0 0 39.0 5.0 241.0 626.75 26.1

Effective Time 11.0 7.0 27.0 25.0 30.0 131.0 41.0 94.0 42.0 65.0 8.0 51.0 531.0 22.1

Results

Drilling Performance. Typical historical drilling times for directional appraisal wells on these fields were 42 to 52 days. Anticipating some improvement in performance, 50/50 budget estimates (a budget estimate based on time, considered as equally likely to be exceeded than beaten) had been prepared with 34 days for a typical moderately deviated well and 46 days for a horizontal section well. Technical limit determination was carried out with the methodology described, arriving at durations of 19.8 days and 27.5 days, respectively.

A comparison between the technical limit, actual, and effective times on Well 1 is presented in Table 4. The fifth column shows where a new technical limit had been set. Furthermore, Well 1 took 22.1 effective days. This was an improvement of 6.6 days compared with effective times of 28.7 and 29.2 days of the previous offset wells; however, Well 1 had 26.2 days of removable time. Table 5 shows the causes of the removable time and the actions taken to prevent recurrence. Most of the removable times were new problems not seen on the offset wells, an indication that the drilling process was still not in control.

TABLE 5POST WELL REMOVABLE TIME ANALYSIS FOR WELL 1 Removable Time (hours) 187.25

Number 1

Problem Back-off in 17 12-in. hole

Cause High drilling torque as a result of formation and aggressive drilling. Excess weight on pipe.

Action and Solution

2 3 4 5 6

Twist off 2 38-in. drillpipe in liner Cleanout of cement in liner Dropped junk basket and nut Correction run in 12 14-in. hole No cement in 9 58-in. shoe

97.00 71.00 64.00 44.00 29.00

7 8 9

Dropped 20-in. casing Rig downtime (total) Problem landing CGB

24.00 22.00 18.00

10

Washout in Monel DC

14.00

11 12 13

Trip in 17 12-in. slow ROP Logging failure Miscellaneous

11.00 4.00 42.50

Reduce drilling parameters. Eliminate string weak-points, review drilling procedures. Procedures. Source stronger pipe. Minimize use. Plug not bumped, no shear Reviewed procedures. Developed inner indication. string method. Nut backed off downhole. Weld nut to shank. Review tool procedures. Right hand bit walk and drop Review bit and BHA selection. Allow left too high. hand lead. Mud syphoning through top Reviewed cement head design. Increased drive when changing over to shoe track to 3 joints. Plan to bump plug release dart. and pressure test casing. Backup tong slipped. Review top drive and casing procedures. Various equipment problems, Review maintenance planning. critical path maintenance. Incorrect bolts-quality Use correct bolts. Review procedures. control. Cuttings build up at wellhead. Stress corrosion cracking. Inspect all monels internal (boroscope). Review quality assurance/quality control procedures. Bit balling. Use PDC bit. Review mud system. Intermittent short in tool Tool returned for repair. Calibrate backup connection. before use. Small events. As available resources allow. 201

SPE Drilling & Completion, September 1998

TABLE 6POST WELL ANALYSIS FOR WELL 3 Starting Technical Limit 10.0 7.5 21.0 25.0 16.5 111.0 36.0 77.0 42.0 63.0 6.0 51.0 466.0 19.4 Actual Time 7.5 6.5 24.5 24.0 19.8 117.0 40.0 103.8 41.3 59.8 9.3 205.3 658.5 27.4 Removable Time 0 0 1.5 4.0 0 1.8 7.5 16.3 3.0 4.8 1.8 155.0 195.5 8.1 Effective Time 7.5 6.5 23.0 20.0 19.8 115.3 32.5 87.5 38.3 55.0 7.5 50.3 463.0 19.3 New Technical Limit 7.5 6.5 21.0 20.0 16.5 111.0 32.5 77.0 38.3 55.0 6.0 50.3 441.5 18.4

Section Drill 36-in. Run 30-in. Drill 26-in. Run 20-in. Run BOP Drill 17 12-in. Run 13 38-in. Drill 12 14-in. Run 9 58-in. Drill 6 12-in. TD logging 4 12-in. liner Total hours Total days

Fig. 6 Comparison of the effective time and removable time wells drilled before the start of the project and Wells 1 and 3. The inefficiency shown is a result of advancing the technical limit and taking a backward view.

Fig. 7Comparison of the effective time and removable time for the comparable portions of the completions. Again inefficiency can be shown by taking a backward view from the final technical limit. 202 SPE Drilling & Completion, September 1998

Well 3 was almost identical to the first well (Well 2 had the horizontal section) and makes a good comparison. The drilling performance for Well 3 is presented in Table 6. Overall performance had improved by 2.8 days, as indicated by the effective time; also, the offshore team was more efficient in many areas, including running blowout preventer and casing. Again, new events happened that had not been anticipated. Only 8.2 days were identified as removable time, 85% of which was accounted for by just two events. An indication that process control was quickly improving. The analysis for Well 2 (with 808 m of horizontal section) was drilled in 40.4 days with 9.4 days of removable time. Adding all wells together, the drilling part of the project took 116.1 days compared to the budget base time of 114 days. Although drilling performance still contained significant removable time, it was a marked improvement on the past (see Fig. 5). A learning curve is evident by comparing effective time and removable time for each of the three wells (Fig. 6). Indeed, it would appear that the operational efficiency had approached the technical limit in just three wells. Such a performance has been described by Brett et al.6 and termed as excellent. Good is the adjective used for approaching the learning curve asymptote in five wells. Completion Performance. Subsea completions technical limits of 10.5 days (new well) and 14.0 days (predrilled well) had been established based on the planning phase described. These compared with 50/50 budget estimates of 16.5 days and 21.5 days, respectively. All these times included production testing. The actual durations achieved averaged 11.3 days for new wells and 15.9 days for existing wells on the six well campaign. The process of refining the technical limit continued as wells were completed. The six completion times are shown in Table 7. The technical limit for a new well had changed from 10.5 days to 8.5 days at the end of the six well campaign. The best new well completion was 9.1 days, which includes all forms of removable time. The subsea completions were finished 33 days ahead of the 50/50 budget times. The completions, by their more mechanically controllable nature, started further along the learning curve than drilling did. The first completion took 25% less time than the 50/50 budget time. Fig. 7 shows the technical limit after the final completion compared with effective time and removable time for each core completion activity. The completions learning curve is still evident in Fig. 7, although less pronounced than with drilling. The drilling and completion campaign not only delivered significant cost savings against the budget, but also delivered subsea wells with productivity 30% greater than expected.
Conclusions

and a technical limit. These concepts were quantified and used in an aggressive method to target the What is possible? question. Technical limit was applied to planning and operations. The following points have been concluded from this approach. 1. Drilling and completion performances can be usefully modeled with the technical limit. 2. The invisible lost time component of this model offers insight into operational inefficiencies that conventional industry lost time systems ignore. 3. Technical limit was used to set the highest performance standards possible. 4. Quantifying and addressing removable lost time provided the maximum opportunity to improve. 5. The technical limit model led to a step change in understanding and performance when applied to three new directional wells and six subsea completions. 6. Adoption of this technique requires courage because it reveals every deviation from the ideal. A no-blame culture was essential to its acceptance.
Acknowledgments

We thank Woodside Offshore Petroleum and its joint venture participants [BHP Petroleum, BP Developments Australia, Chevron Asiatic, Japan Australia LNG (MIMI), and Shell Development Australia] for support throughout this project. Additional thanks to Sedco Forex, Baker I.S., and all personnel that contributed to the development and application of the technical limit method. Personal thanks to Phillis Harley for tireless support from the beginning of the project and through the preparation of this paper.
References
1. Noerager, J.A., White, J.P., and Floetia, A.: Drilling Time Predictions From Statistical Analysis paper SPE/IADC 16164 presented at the 1987 SPE/IADC Drilling Conference, New Orleans, Louisiana, 1518 March. 2. Shute, J. and Alldredge, G.: Conoco Cuts North Sea Drilling Time by 40%. World Oil (July 1982). 3. Huber, D.D. and Walton, H.: A Realistic Goal From a Semi: Drill to 10,000 ft in 10 Days, World Oil (September 1983). 4. Walters, N.S.: Maximizing Well Potential: An Integrated Approach, paper SPE 36997 presented at the 1996 SPE Asia Pacific Oil and Gas Conference, Adelaide, Australia, 2831 October. 5. Kadaster, A.G., Townsend, C.W., and Albaugh, E.K.: Drilling Time Analysis: A Total Quality Management Tool for Drilling in the 1990s, paper SPE 24559 presented at the 1992 SPE Annual Technical Conference and Exhibition, Washington, D.C., 47 October. 6. Brett, J.F. and Millhiem, K.K.: The Drilling Performance Curve: A Yardstick for Judging Drilling Performance, paper SPE 15362 presented at the 1996 SPE Annual Technical Conference and Exhibition, New Orleans, Louisiana, 58 October.

The results presented here could be considered lucky with such a small data set; this is not disputed. The important conclusion with the methodology outlined lies with a focus on the theoretical well

SI Metric Conversion Factors

ft 3.048* in. 2.54

E01 m E00 cm SPEDC

TABLE 7FULL COMPLETION TIMES Technical Limit (days) 11.4 10.7 10.9 8.9 14.9* 13.2* Actual Time (days) 12.5 12.8 10.8 9.1 19.3 14.2 Removable Time (days) 1.4 2.3 1.2 1.2 4.7 2.0 Effective Time (days) 11.1 10.5 9.6 7.9 14.6 12.4 David F. Bond is a principal drilling engineer for Woodside Energy Ltd. in Perth, Australia, responsible for technical integrity and performance of all Woodside drilling activities. He holds a BS degree in mechanical engineering from the U. of Leeds, U.K. Phil W. Scott is an implementation manager for new systems being developed by Woodside Energy Ltd. in Perth, Australia. His recent focus has been on human elements in well construction. He holds a BS degree in civil engineering from the U. of Western Australia. Peter E. Page is a senior completions engineer for Woodside Energy Ltd. in Perth, Australia. Previously, Peter worked for Shell Intl. He has worked for Shell in the U.K. and in Norway before moving to Australia in 1992. Tom M. Windham is a drilling superintendent with Chevron Overseas Petroleum Inc., in Escravos, Nigeria. He was seconded to Woodside from November 1993 to December 1996. He holds a BS degree in petroleum engineering from West Virginia U.

Completion 1 2 3 4 5 6

* Include running additional equipment.

SPE Drilling & Completion, September 1998

203