Drilling Technology

Stringent drill-in fluid design results in prolific deepwater horizontal well

BP’s horizontal, openhole gravel pack in Marlin field development Well A-4 in the Gulf of Mexico was successfully completed with fluids specifically designed and tested to optimize drill-in and well clean-up

Stu W. Gosch, BP; Dale W. Bradford, BHP Billiton Petroleum; Brian A. Butler, Mike F. Adkins and Joe R. Murphey, TETRA Technologies, Inc.

his article overviews the drill-in and completion of BP-operated offshore development well Marlin A-4 in the Eastern Gulf of Mexico. Objectives of the well are outlined. Design and selection of a drill-in fluid to meet selected inhibition, toxicity, rheology and filter-cake cleanup is described, which led to use of TETRA Technologies’ proprietary, brine-based PayZone drilling fluid. Drilling and completion operations are described, which led to a successful gravel pack and sustained production of 87 MMscfd gas and 9,700 bpd condensate.

Conclusions from this project are that, through use of a comprehensive testing program, fluids considered optimal over a wide range of operating conditions can be developed. Close adherence to parameters established in the design phase of Well A-4 led to successfully drilling, completing and producing the well. All phases of the operation went smoothly and all objectives for the well were met or exceeded. Productivity results are considered outstanding, as the subject well is one of BP’s most successful wells in the Gulf of Mexico, and one of its highest net revenue generators.

With proper design, brine-based fluids can be made very inhibitive and are quite suitable for shaley environments. Efficient openhole, gravel-packed completions are readily achievable in these environments, as evidenced by the subject well and a growing database of similar applications.

Introduction

A horizontal, openhole gravel pack was the well configuration of choice for BP’s Marlin A-4 development well, located in Viosca Knoll Blocks 871 and 915 in the Eastern Gulf of Mexico in 3,230 ft of water. The drill-in fluid design was considered a critical factor influencing overall well and project success.

Given the importance of the subject well and the peculiarities involved in this deepwater project, a comprehensive study for the drill-in fluid and well cleanup design was undertaken. Several design parameters were driven by project economics, while others were a function of operational limitations. Rigorous testing resulted in what proved to be an optimal fluid system that met the complex set of design criteria established to drill, complete and produce the well efficiently.

The well’s design phase proved successful and an optimal fluid was developed. The well was successfully drilled and completed and production results are considered outstanding. An openhole gravel pack was executed as designed, resulting in complete pack placement and a highly efficient completion. The investment in this comprehensive study is easily justified, as Well A-4 is considered one of BP’s most successful wells in the Gulf.

Plans called for Well A-4 to be completed near the gas-oil contact in the main pay sand. The reservoir is a Middle Miocene, deep-sea, rich-fan deposition, with moderate sorting and very limited cementation. Average gross thickness is 80 – 120 ft, with a typical net-to-gross ratio of 70% to 80%. The majority of the field’s reserves (80%) are contained in the Main Pay Sand.

Well A-4 Objectives

Initial plans called for an interval length of about 2,500 ft. However, after geologic reinterpretation and inflow modeling work, it was decided that – with an efficient completion – an interval length of about 800 ft would allow production targets to be met.

Three major objectives were established for Well A-4. With the openhole configuration, all three objectives would be largely controlled by performance of the drill-in fluid and subsequent cleanup success. The major objectives were:

1. Accomplish an extremely stable wellbore to allow efficient placement of a long-duration, openhole gravel pack.
2. Accomplish a thorough cleanup of the entire productive interval – considered a requirement for an efficient, high-rate well completion.
3. Achieve a durable, long-lived well, capable of producing all available reserves without need for intervention.

Drill-In Fluid Design/Selection

For Well A-4 to be fully successful, the drill-in fluid would have to meet several important criteria. The fluid ultimately selected would have to demonstrate extremely good performance in shale inhibition, marine toxicity, fluid rheology and efficient filter cake cleanup.

Additionally, the fluid would have to meet a 4/30 PCT, i.e., 4,000 psi/30°F pressure crystallization temperature, requirement. Consequently, a robust test matrix was undertaken for the purpose of optimizing the drill-in fluid and well-cleanup treatment. The fluid ultimately selected for the project was a proprietary PayZone drilling fluid made from a mixed sodium-calcium base brine. This proved to be the best solution for this project, as it met or exceeded all prescribed performance criteria.

Fluid crystallization. Inadvertent crystallization of salt from brine and the resultant problems are well known. This phenomenon is complicated in deepwater environments where fluids can be placed under substantial pressures at seabed conditions. Discovery of this phenomenon led to development of clear fluids with PCT ratings, rather than traditional TCT (true crystallization temperature) ratings. Since Well A-4 is located in water deeper than 3,200 ft and mudline temperature is 38°F, potential crystallization of brines at seabed conditions was a major concern.

Proper fluid design for deepwater applications includes projecting the maximum pressure to be exerted on the fluid at the seafloor. This parameter for Well A-4 was set at 4,000 psi. With added safety factor, all brine fluids used during the drill-in / completion operation were required to meet or exceed the 4/30 PCT spec, noted above.

Designing for PCT purposes had to be weighed against the marine-toxicity and shale-inhibition characteristics of the fluid. Testing was done on brines composed of NaBr, CaBr₂ and combinations of sodium and calcium salts. Densities were then increased at 0.5-ppg increments and reevaluated. Various ratios of sodium and calcium were evaluated to optimize the brine composition. Work next centered on selecting the formulation that exhibited the best performance in the other areas studied.

Density. The final drilling-fluid density required was not finalized until shortly before project startup. Modeling work conducted by BP suggested that a density range of 12.0 to 13.0 ppg would be required. In this application, fluid density was not governed by reservoir pore pressure (10.3 ppg EMW), but rather by borehole-stability requirements. Ultimately, the decision was made to begin drilling with 12.0-ppg fluid, and weight up – as dictated by well behavior – to as high as 13.0 ppg. Given the density range established, it was necessary to examine the best ways to accomplish the density increases, if the need arose.

After evaluating anhydrous calcium bromide, calcium carbonate and calcium bromide spike fluid, anhydrous calcium bromide became the clear choice as the best weight-up method for this project. It was determined to be most cost-effective and would not require large quantities to be stored on location. Pilot testing revealed that there would be very little effect on the fluid’s rheological properties, and volume increases would be minimal.

Shale inhibition. With the uncertain geology, it was possible that the well path would intersect fairly long shale sections. As drilling reactive shale can be a major factor influencing design / performance of brine-based fluids, it was decided to plan for this contingency. This required that a fluid with maximum inhibitive qualities be developed.

Good shale inhibition is obviously very important for drilling success. However, less obvious is the importance of this property for production optimization purposes. Poor inhibition may not only cause borehole stability problems, it can result in undue contamination of the drilling fluid with native (insoluble) drill solids. This contamination issue ultimately causes a dramatic reduction in filter-cake solubility and degradability. This, in turn, creates a situation in which filter cake and entrained solids become a serious production-impairment mechanism that is difficult or impossible to treat out effectively.

A large portion of this study concentrated on developing the most inhibitive formulation possible. It was decided to use a shale-dispersion test for the purpose of assessing inhibition. A total of 19 different base fluids were tested, including sodium and calcium brines, as well as multi-salt combinations, see table. As part of the test matrix, the various fluids were evaluated, with and without inclusion of PayZone StrataFix (a glycol additive) and cesium formate, both considered potential shale-stabilizing agents.

Addition of the glycol dramatically increased the fluids’ inhibitive qualities in shale-dispersion testing. The best fluids tested yielded 85% to 95% recovery in the tests, see Fig. 1. Results in this range are considered very good, and fluids exhibiting this behavior in lab tests have proven to be inhibitive in field use. The most inhibitive systems tested contained both cesium formate and the glycol-shale-stabilizing additive. In comparing brine compositions, the sodium and mixed brine systems performed best. The calcium bromide system yielded a poor recovery of only 55%. This poor result was somewhat unexpected, due to calcium bromide’s widespread use in the Gulf of Mexico.

Fig. 1. Effect of StrataFix glycol on shale dispersion

The glycol had no negative impact on fluid properties, and had the added benefit of enhancing fluid lubricity. For these reasons, glycol was considered an important and necessary additive for this project. Cesium formate also appeared to be beneficial, but was eventually excluded due to complications involving cost and marine toxicity.

Marine toxicity. Space limitations on the tension-leg platform made it impractical to capture the drill cuttings from the well. Because of this, a design parameter was established such that the ability to discharge cuttings and fluid would be maintained throughout the drilling operation. EPA compliance requires a minimum passing LC₅₀ of 30,000 ppm using established test protocols. However, as insurance against a failing test, BP required that the drilling fluid exhibit an LC₅₀ of 100,000 ppm or higher. As this was considered untypical for traditional brine-based fluids in this density range, a large amount of formulation work was required for this purpose.

Fluid optimization involved balancing toxicity results against the numerous other aspects of the fluid. After a large number of iterations and comprehensive additive screening, the drill-in fluid ultimately selected exhibited an LC₅₀ of about 185,000 ppm. This was well above the minimum set by BP, and considered very high, as compared to conventional calcium-based, low-solids fluids of this density.

Rheology. A great deal of planning went into optimizing the drilling fluid rheology. To avoid possibly exceeding formation fracture pressure, a ceiling of 13.2-ppg ECD was imposed. Given that this well was nearly horizontal, hole cleaning was obviously a major concern. Fine adjustments to fluid formulations were made and extensive hydraulics modeling was conducted.

Both Power Law and Herschel-Buckley models were used in hydraulics modeling and each gave considerably different results. The Herschel-Buckley model consistently predicted higher friction pressures than did the Power Law model. Due to very good historical matching between field PWD data and a Power Law model, it was opted to use this for ECD predictions. In working through this process, the most suitable fluid formulation was developed.

Filter-Cake Cleanup/ Return Permeability

Three main criteria were established for clean up of the productive interval. The aim was to remove any production impairment mechanisms stemming from the drill-in process. These criteria were:

1. Contact any and all damage present, i.e., accomplish full heel-to-toe cleanup of the interval.
2. Accomplish efficient flowback at worst-case (very low) inflow velocities.
3. Accomplish the above objectives using a non-acid clean-up process due to concerns with acidization.

Breaker studies were conducted using a dynamic (stirred) modified HTHP fluid-loss cell. To make the tests as realistic as possible, filter cakes were built, using up to 30 ppb of Marlin shale as a contaminant. Several different breaker concentrations were evaluated before the desired break time was attained. The target break time for Well A-4 was set at 12 hr to allow time (with safety factor) for gravel-pack placement.

A series of tests were also conducted to assure fluid-fluid and fluid-rock compatibility. Various combinations of reservoir crude, drill-in fluid, gravel-pack carrier fluid and acids were tested for sludging, emulsion and precipitation tendencies. Core plugs were injected with drill-in fluid filtrate, gravel-pack carrier fluid and completion brine, and return permeability was measured. No significant permeability impairment or incompatibilities were observed with any of the fluids tested.

Drilling Operations

As a result of a "right-scoping" exercise, the original well plan was revised to reduce the lateral length to about 800 ft, vs. the originally planned 2,500 ft. An 806-ft (MD) interval was drilled, and TD was called at 13,390-ft MD after drilling about 100 ft of shale below the base of the sand. Of the 806-ft interval, 730 ft were completed. Log analysis indicates that, of the completed interval, about 498 ft is considered productive reservoir.

Drilling operations went without incident. Hole behavior was considered excellent throughout the operation. While drilling, actual ECDs were within 0.05 ppg of those predicted by the model. During drilling, the amount of drill solids being entrained in the fluid was closely monitored. This was done at least once per tour via a gravimetric test to measure the acid-insoluble materials.

As part of pre-job planning, a limit of 3% (by volume) of acid-insolubles was established. Although the drilled interval contained a fair amount of shale and dirty, poor-quality sand, fluid cleanliness did not become an issue. With the optimized fluid, the level of acid-insoluble materials in the fluid reached a maximum level of 1.12% – an indicator of the fluid’s inhibitive qualities.

Completion Operations

After displacing the well to clean completion fluid, a gravel-pack assembly with 4-in. base pipe, prepack screen was run to a depth of 13,314 ft, with the shoe positioned just below the bottom of the sand. Screens went smoothly to bottom without circulation or manipulation. A circulating gravel pack was then performed with 20/40 sand. The pack was pumped at 2.0 bpm, with a sand concentration of 1 ppg. Extremely good return rates (>95%) were observed throughout the pumping of the alpha wave, with only slightly lower returns during beta wave placement.

Pack placement went according to schedule, with just over 15,000 lb of sand placed below the crossover tool. Based on volume of sand placed and pressure profiles seen, it is evident that complete packing was accomplished, and the wellbore was drilled to gauge.

Simultaneous cleanup / gravel pack. A PayZone ACT breaker system was shipped to location as a liquid concentrate. The breaker presented no serious handling or personnel exposure issues, and was easily added to the existing completion fluid on location. This fluid was, in turn, used to make-up the gravel slurry, and the job was pumped as usual.

This process proved successful, and no acid treatment was used to break the filter cake, or to dissolve the calcium-carbonate bridging solids. Drill solids in the fluid were controlled during drilling and proved non-problematic to the clean-up operation.

Production results. The initial production target for this well was about 80 – 90 MMscfd. The A-4 completion was successful in meeting this target, even though interval length was shortened to 800 ft. Ten months after first production, A-4 was still producing about 87 MMscfd gas and 9,700 bpd condensate, with 1,375 psig FTP.

Downhole-pressure-gauge data indicates total drawdown at 650 to 750 psi. Pressure transient analysis has been influenced by phase segregation and wellbore storage effects and is, so far, considered inconclusive.

Notes

Technical paper (AADE-02-DFWM-HO-03) outlining this case study was presented at the AADE 2002 Annual Technical Conference. For more information on this project, the paper can be accessed at www.payzone.com.

PayZone, StrataFix and StrataFix ACT are registered trademarks of TETRA Technologies, Inc.