Mark Stewart
Schlumberger Dowell
Bakersfield, Calif.
Dave Stewart
Chevron U.S.A. Production Co.
Houston
Mike Gaona
Chevron U.S.A. Production Co.
Bakersfield, Calif.
An aggressive development program and advances in fracturing design and equipment have allowed a once declining field to more than double production since 1989.
In the Lost Hills field, the alliance between Chevron U.S.A. Production Co. and Schlumberger Dowell has made significant progress towards the objectives of improving fracturing efficiency, reducing cost, and improving well productivity. The results have increased development well profitability.
FIELD HISTORY
About 2 billion bbl of oil-in-place are present in the massive diatomite deposits of California's Lost Hills field, about 45 miles northwest of Bakersfield, Calif. (Fig. 1). Massive hydraulic fracturing treatments, 2,5003,000 lb of proppant/net perforated ft, are an integral part of developing these reserves.
An exclusive fracturing alliance initiated in 1990 between Chevron U.S.A. and Schlumberger Dowell has improved profitability of the Lost Hills field. Significant well productivity increases and cost reductions have resulted from an aggressive field development program and focused process improvement efforts by the alliance team.
As of July 1, 1994, over 2,000 fracture stages have been performed during the completion of over 600 wells, placing in excess of 1 billion lb of fracture proppant. These efforts have contributed to the 250% production increase from a pre 1989 level of 6,000 bo/d to the current 15,000+ bo/d. Fracturging costs have been reduced by 40% relative to 1988.
The Lost Hills field was discovered in July 1910. Initial development targeted the Tulare (Pleistocene), and Etchegoin (Pliocene) formations. Standard Oil Co. of California (Chevron U.S.A.) tested deeper intervals in late July, 1913. Initial production from the Miocene Monteren, formation was 60 bo/d and 60 bo/d through perforations from 3,100 to 4,500 ft. Well completions in the first half of the century were either open hole or slotted liner. Jet perforating began in 1945.
Relatively few wells were drilled and completed in the Monterey formation before 1948, when development activity increased. In 1948, Chevron U.S.A.'s delineation well-Monte Cristo 181X-flowed 480 bo/d, 680 Mcfd, and 100 bw/d from the naturally fractured Antelope member of the Monterey formation. By 1982, favorable production had encouraged operators to increase the number of wells completed in the Monterey formation to 211 extending the field over 2 miles to the southeast.
GEOLOGY
The Lost Hills field is part of a northwest-southeast striking anticlinal trend that extends 50 miles northwestward to the Coalinga anticline. The Lost Hills anticline is asymmetric, and is about 1 mile wide and 12 miles long (Fig. 2). The anticline is oriented northwest southeast, parallel with the San Andreas fault and roughly perpendicular to regional compression.
During the late Miocene (5-11 million years ago), a portion of the present day San Joaquin Valley was a restricted marine basin. Cool temperatures and upwelling currents allowed diatom populations to flourish near the surface. Diatoms are single-celled microscopic plants with porous, silica exoskeletons.
Diatoms and various amounts of fine-grained sand, silt, and clay accumulated under low-oxygen conditions to form the laterally continuous deposits of the Monterey formation. Four different members of the Monterey formation, in ascending order, are present in the Lost Hills field. These are the McDonald shale, Antelope shale, Brown shale, and Belridge diatomite (Fig. 3). The lithology ranges from siliceous shale (chert) in the McDonald member, to diatomite in the Belridge diatomite member.
Although porosity of the Belridge diatomite reservoir is generally very high (over 60% in the cleaner diatomite), the permeability is very low, generally less than 1 md. Petrophvsical changes with depth alter silica in the Opal-A phase to the Opal-CT phase. With further diagenesis the Opal-CT phase silica is altered to the Quartz phase (chert). All three silica phase transitions control hydrocarbon accumulation, acting in combination with the anticlinal structure and stratigraphy to create a number of separate reservoirs (Fig. 3).
The Belridge diatomite does not exhibit an open, well-developed interconnecting natural fracture-matrix network. The natural fractures that do exist are present mainly at the Opal A-CT transition, with alignment normal to the axis of the Lost Hills anticline. Table 1 lists average rock and fluid properties for the Belridge diatomite.
BEFORE 1987
At Lost Hills, hydraulic fracture stimulation started in the 1960s. Early attempts targeted the Antilope shale member of the Monterey formation. Treatments were at low injection rates, 20 bbl/min, with saltwater carrying fluid and 10-20 mesh silica sand or 8-12 mesh walnut hulls as proppant. These completions performed poorly, and as a result, few fracture stimulations were performed until the mid 1970s. In mid-1970, fracturing increased because of cross linked hydroxpropyl guar (HPG) and guar carrier fluids, and stair-stepped proppant concentrations (Table 2), ranging from 3 to 10 lb of proppant added/gal(PPA).
Performance was mediocre, however, because of small proppant volumes, 500-1,000 lb/ft, and inefficient fracture fluids.
Initial production response averaged between 20 and 25 bo/d. These stimulations generated short fracture half-wing lengths of less than 75 ft and yielded poor effective propped heights because of excessive pad volumes.
1987-1990
Beginning in 1987, Chevron accelerated primary development of the Befridge diatomite reservoir. As part of that effort, Chevron initiated massive hydraulic fracture stimulations, 3,000 lb/ft. Slurry injection rates increased to 50 bbl/min with maximum proppant concentrations of 16 PPA. Gel loadings were usually 40 lb/1,000 gal (PPT). Several tanks of fracturing fluids were batch mixed on site, giving rise to logistical and quality control problems.
Well productivity increased significantly when compared to the earlier type of stimulations, but costs were high and the jobs were not as profitable as Chevron had anticipated. Because of the magnitude and complexity of the Lost Hills development program, Chevron be,-an the search for a technical partner to jointly attack these issues. In late 1990, Chevron and Schlumberger Dowell joined into an exclusive fracture stimulation services alliance.
AFTER 1990
During 1986-1988, Chevron evaluated three service companies based on service, price, and well performance and thus established a framework for one of the first pumping alliances.
In 1990, Chevron U.S.A.'s western business unit and Dowell established a 5-year fracturing alliance agreement. To achieve the mission statement of refining fracturing quality and well performance, both companies agreed to a team concept that allowed for maximizing profits while setting the highest standards for both safety and environmental practices. The three initial objectives for this alliance were to:
- Make significant improvements in joint profitability
- improve service quality through reduction of product variability
- Provide fracturing services with on site project management of service personnel, well services, and engineering.
The primary goal of maximizing profitability through quality improvements to the fracturing process was achieved by the following:
Meeting annual fracturing requirements for both new and remedial wells through improved planning
- Establishing a performance metrics analysis system that measured fracture treatment cost and effectiveness
- Improving field/office communication through office sharing of both service company and operator personnel
- Establishing a fair and equitable fracturing services compensation package based on performance and market forces
- Implementing a service quality improvement program that measures equipment, personnel, safety, and environmental performance
- Developing annual budgets based on forecasts, manpower commitments, and expenditure levels that incorporate the financial goals of both Chevron and Schlumberger Dowell
- Establishing mutually agreed upon environmental, safety fire, and health goals 0 Forming a technical team to continuously evaluate and improve the fracturing process.
Early analysis of previous stimulation designs led to the conclusion that merely pumping increased proppant volumes would not optimize the fracturing program. Listed in Table 3 are major areas that underwent intense evaluation and change.
FLUID
Initially, two base gels were used at Lost Hills: hydroxypropyl ouar (HPG) and guar. Tubing treatments experiencing high-shear conditions used primarily HPG while jobs down the casing used guar gels.
Original gel loadings for Lost Hills treatments were 40-50 lb/1,000 gal. Several studies have shown that reducing polymer loads and consequential the amount of potentially damaging materials improves fracture conductivity. Trade-offs encountered by reducing polymer loadings are increased fluid leakoff and decreased proppant carnning ability.
Field experimentation and improved equipment have yielded better viscosity control that has reduced gel loadings to 25-30 lb/1,000 gal. Further gel loading reductions will probably not be attempted because cross link quality is affected by an insufficient number of negatively charged crosslink sites
Liquid gel concentrates (LGC) are currently included in all Lost Hills stimulation treatments. Prior to LGC, all treatments were batch mixed hours or even days before the actual treatment. Advantages of continuously mixed gels include:
- Elimination of time-consuming mixing of frac tanks
- Consistent gel viscosity free of "fish eyes" Gel-loading ramping flexibility
- Reduced costs
- Enhanced post job tank cleaning
- Elimination of expensive disposal of tank bottoms.
The change-over to LGC and improved fluid mixing technology allowed fracture fluid pad sizes to be reduced by over 4017, since 1988,
At Lost Hills, most fracture stimulations involve shallow (less than 3,000 ft) low temperature (90-150 F.) completion intervals. For this reason, borate crosslinkers have been used exclusively because of their shear healing ability, lower cost, and minimal damage compared to metallic crosslinkers. Also, persulfate/amine breaker systems have improved cleanup and conductivity when compared to conventional enzyme breaker systems.
To control clay swelling in the diatomite, a base salt must be added to the mix water prior to treatment. Initial studies indicated that 3% potassium chloride (KCI) was appropriate as this concentration matched the original formation water salinity, averaging 10,000-15,000 PPM
Recent use of 2% KCI showed no loss in well productivity; therefore, the base fluid was changed to 2% KCI in 1994.
Additive usage during the mid-1980s included surfactant, clan, stabilizers, scale inhibitors, defoamers, bactericides, breakers, buffers, crosslinkers, activators, and KCI. Profitability-driven process improvements have eliminated all surfactant, clay stabilizers, and most scale inhibitors, as well as reduced select additives. Long-term production has not been adversely impacted by the elimination or adjustment of these additives.
PREAPPOINTS
Early stimulation field trials using, 20/40, 12/20, and resin-coated preappoints showed very little difference in overall well productivity. Because of the relatively fracture closure stresses (less than 2,000 psi), proppant crushing has not been a problem. Like wise, well productivity has not been greatly enhanced by larger mesh preappoints, such as 12-20, capable of higher conductivity.
Because of the extremely soft nature of the diatomite reservoir rock, less than 200,000 Young's modulus, it had been thought for several years that creep or embedment caused the rapid production decline. Embedment is the process by which proppant is pushed into the softer reservoir rock, reducing fracture width proportional to the grain diameter embedded in the reservoir rock.
Tests conducted at Chevron's La Habra research center indicated that embedment of 0.5 grain diameter resulted with 20/40 mesh sand at 2 psf fracture-pack densities and 1,500 psi closure pressure. Therefore, embedment into the diatomite can be described as minimal, and is not a concern because of the actual frac pack densities of 5-6 psf.
Creep is the process whereby formation or reservoir fines migrate into the newly created fracture pack over time. Laboratory testing for creep also showed negligible reductions in fracture conductivity.
Prior to the development of process controlled blenders, increasing concentrations (generally 1-3 PPA) of proppant were added in a stair-step manner. Today, new equipment achieves aggressive proppant ramping beginning at 6 PPA through 16 PPA. This allows sand ramping as a continuous function. In addition to improved fracture conductivity, the smoother proppant ramp helps prevent near well bore screen outs.
Proppant concentrations currently are held to a maximum of 16 PPA. This concentration is less than the typical 1987 vintage fracs that routinely, tailed-in with 18 ppg added. Due to the increasing number of nearwell bore screenouts at 18 PPA and blender throughput limitations, it was decided to pump a maximum of 16 PPA. This small reduction helped eliminate the nearwell bore screenout problem.
In current frac designs, over 50% of the entire treatment is at the maximum proppant concentration of 16 PrA. Elevated proppant concentrations such as these yield simulated fracture conductivities of 800-1,500 md/ft.
Proppant volumes are measured on the basis of pounds placed per foot of perforated interval. Proppant volumes increased from 500 lb/ft to a maximum of 3,000 lb/ft during the 1989-91 period. Currently, all wells are stimulated with 2,500 lb/ft.
Based on radioactive tracer evaluation, perforated intervals have been reduced from the original maximum of 300 net ft/stage to the current maximum of 250 net ft/stage. Total proppant volumes typically average 2.53.0 million lb for new well producer completions with four to six frac stages.
ENGINEERING EVALUATION
Data fracs performed on the McDonald, Antelope, Brown shale, and Belridge diatomite intervals have obtained closure pressure, fluid efficiency, and through computer simulation of pressure data, a frac geometry model.
Closure pressure is critical because nearly all fracture modeling equations are based on this parameter. Fluid efficiency is also important in the determination of correct pad volumes. If excess pad volume is pumped, job cost and well cleanup time are increased. Current fluid efficiencies in the diatomite interval are in the 50-70% range.
Current frac treatments are designed from empirical results and simulations. The treatments are high volume, 2,500 lb/ft, and high rate, 50 bbl/min casino and 35 bbl/min tubing. Aggressive proppant ramping from 6 to 16 PPA is achieved with continuously med guar polymers using state-of-the art process-controlled blenders, gel mixers, and computerized monitoring vans.
By operating two to three frac fleets throughout the year, Schlumberger Dowell has met the alliance team's yearly fracturing objective (Table 4). Improvements in stimulation technology, well selection, and completion interval identification have resulted in significant well productivity improvements over 1970 vintage treatments (Fig. 4).
EXECUTION
Job execution has been an important focus for the alliance team. One unique improvement was the implementation of central fracturing sites in 1992. This allows fracturing of multiple wells from fixed, centrally located sites (Fig. 5) instead of relocating all equipment for each well.
Wells have been routinely fractured at distances in excess of 3,000 ft from the central fracturing sites. Because of the relatively tight well spacing of 2-1/2 acres and improved planning, up to 40 wells (100 plus stages) have been fractured without relocating the fracturing equipment.0
The central sites have lowered costs by reducing personnel requirements, well completion cycle time, and the number of mobilizations and demobilizations. Additional benefits include improved environmental management and safety control.
Since implementation in late 1992, the central site strategy has been used to stimulate over 100 wells, placing about 200 million lb of proppant.
In May 1993, alliance concerns about the safety and efficiency of surface fracturing service tubulars prompted design and manufacture of a new 5.5-in. OD surface fracturing line. This new pipe replaced three smaller 3.0 in. OD lines. Fig. 6 shows the 5.5-in. line laid to a distant well site.
The larger internal diameter reduces slurry velocities, thereby decreasing internal pipe wall erosion and increasing service life, This new surface fracturing pipe was also designed to reduce installation time.
The design included (WECO style) hammer union connectors and 45-ft joint lengths to reduce make-up time. More than 3,000 ft of 3.5 in. OD, 17 lb/ft N-80 fracturing line was manufactured. In addition, a specialty modified side-loading fork-lift truck was designed and built, further reducing the time and cost associated with moving and installing the line.
These improvements, along with previously mentioned fracturing design changes, have reduced overall fracturing cost by 40% since 1988 (Fig. 7). Recent process improvements, which have not been fully evaluated to date, along with future improvements should ensure optimization of profitability.
The alliance team also focused on improving coordination of drilling and completion schedules. Annual drilling and completion plans are finalized by September of the previous year. During these planning sessions, central fracturing locations are designated based on well location, drilling order, and completion sequence.
A service radius of about 3,000 ft is used to draw site boundaries. Typically, much of the annual fracturing program of over 400 fracturing stages can be completed from no more than five central fracturing sites.
The drilling and completion schedules are carefully planned to eliminate surface location conflicts between drilling and completion rigs. This is important because of the aggressive development schedule, tight well spacing, and large amount of service activity.
At a more detailed level, a 3-week fracturing schedule is prepared or updated each Thursday. This schedule is used by all contractors (rig contractor, wire line companies, tool company etc.) to plan their activities.
The 3-week schedule was originally developed to ensure a continuous supply of proppant from sand quarries located near Kasota, Minn. Rail delivery time from Minnesota to Bakersfield, Calif. ' is about 3 weeks, and the 3 week schedule ensured timely delivery of proppant.
The schedule later developed into a comprehensive short-term planning, tool for all customers and suppliers involved with the completion process.
ACKNOWLEDGMENTS
The authors would like to thank management from both Chevron U.S.A. and Schlumberger Dowell for allowing the publication of this article.
Copyright 1994 Oil & Gas Journal. All Rights Reserved.
Issue date: 11/21/94
