The first acrylic acid was created in 1843. Methacrylic acid, derived from acrylic acid, was formulated in 1865. The German chemists Fittig and Paul discovered in 1877 the polymerization pro…Full description
heat of solutionFull description
acrylic acid summary of information
Acid s;urry report
SPE Comprehensive ExamFull description
Sebuah makalah tentang mikrohidroDeskripsi lengkap
OVERVIEW For this month’s issue of featured well-stimulation papers, the ongoing work in acidizing earns my recognition as the most enlightening and yet overdue progress. Although the papers spotlighted in this issue are strong presentations on diverse subjects, I was most surprised by the volume of engaging papers focusing on acids and fracture acidizing. There are great papers in my list with information on horizontal wells, gravel packing, and viscoelastic fluids; however, each of these technologies is young compared to acidizing reservoir rock. In one of the additional-reading selections, “Fracture Acidizing: History, Present State, and Future,” we learn that acidizing carbonates dates back to 1895. Therefore, the use of acids to increase production is more than a century old. With Wit h many many of of the the moder modern n acidi acidizing zing theo theorie riess develo developed ped arou around nd 1972 1972 by Neir Neirode ode and others, it is still a 35-year-old technology that needs a fresh look with modern methods. Whether we are studying etching behavior to forecast conductivity or predicting long-term reductions in performance caused by creep, we still have a lot to learn about this venerable practice. Therefore, the question becomes: “How can acidizing oil wells be more than 100 years old, yet we just now are beginning to unravel these fundamental concepts?” The world produces approximately 85 million BOPD, and assuming just USD USD 50/B, 50/B, we generat generatee more than USD USD 1.5 1.5 tril trillio lion/yr n/yr in in world world oil revenue. revenue. Whatever What ever the the historica historicall disconnec disconnectt has been between between oil revenu revenues es and fundin fundingg oil science, I am pleased to recommend these compelling technical papers on subjects vital to our industry, including the common practice of pumping acid JPT into a formation to increase production. Charles Hager, SPE, is a senior consultant with NSI Technologies. For the last 17 years,, he years he has has focus focused ed on on applyi applying ng and inte integrat grating ing the scien sciences ces of hydr hydraulic aulic-fra -fractur cturing ing analysis and modeling, pressure-transient analysis, and reservoir simulation. Hager earned a BS degree in petroleum engineering from the University of Alabama, and he serves on the JPT the JPT Editorial Committee.
Well Stimulation additional reading available at the SPE eLibrary: www.spe.org SPE 108075
“Horizontal-Well Completion and Stimulation Techniques—A Review With Emphasis on LowPermeability Permeabili ty Carbonates” by Valdo Ferreira Rodrigues, SPE, Petroleo Brasileiro, et al. SPE 107772
“The Effects of Acid Contact Time and the Resulting Weakening of the Rock Surfaces on Acid-Fracture Conductivity” by M.G. Melendez, Texas A&M University, et al. SPE 107760
“Acid Stimulation of Extended-Reach Extended-Reac h Wells: Lessons Learnt From N’Kossa Field” by J.M. Mazel, Total, et al. SPE 106371
“Fracture Acidizing: History, Present State, and Future” by Leonard J. Kalfayan, SPE, BJ Services
JPT • JULY 2007
WE LL ST IMUL AT ION
A Small-Scale Fracture-Conductivity Study
A series of acid-fracture-conductivity tests was conducted that simulated flow in a hydraulic fracture, both in the main flow direction along the fracture and in the fluid-loss direction. Three commonly used acid-fracturing fluids were tested at 200 and 275°F. The acid-fractureconductivity apparatus is similar to a standard American Petroleum Institute (API) fracture-conductivity cell, but with the ability to hold core samples that are 3 in. thick in the leakoff direction.
Success of the acid-fracturing process depends on the resulting fracture conductivity, which is difficult to predict because it depends on a stochastic process and is affected by a wide range of parameters. Most predictions of conductivity are made with the empirical correlation developed by Nierode and Kruk. This correlation was based on experiments with 1-in.-diameter, 2- to 3-in.-long fractured cores with no fluid loss through the rock samples. To ensure that laboratory experiments represent field conditions, the phenomena that occur in the acid-fracturing process must be scaled properly. In this study, experimental conditions were scaled to give the same magnitude of acid transport along the fracture, acid leakoff, and acid reaction at the fracture face as occurs in field treatments.
Introduction Acid fracturing, a well-stimulation process in which acid dissolves reservoir rock along the face of the hydraulically induced fracture, is expected to create lasting conductivity after fracture closure. However, conductivity after fracture closure requires that the fracture face be nonuniformly etched by the acid while the strength of the rock is Acids maintained at high levels to withstand Three acid systems were used in the closure stress. At low closure stress, conductivity tests: a gelled-acid systhe etched pattern of the fracture face tem with an acid-soluble polymer as a should have a dominant influence on gelling agent, an acid-in-oil emulsion, the resulting fracture conductivity as and an acid/viscoelastic-surfactant acid long as the rock strength can with- solution. The polymer-gelled-acid sysstand the load. As the closure stress is tem contained 15% HCl, 2.5% gellincreased, surface features along the ing agent, and a corrosion inhibitor. fracture faces may be crushed, which The emulsified acid was a mixture of makes fracture conductivity more 28% HCl and diesel with an emulsidependent on the rock strength than fier and a corrosion inhibitor. The on the initial etching pattern. diesel forms the external phase of the emulsion. The viscoelastic-surfactant This article, written by Assistant Tech- acid has unique “self-diverting” charnology Editor Karen Bybee, contains acteristics. The fluid viscosity increases highlights of paper SPE 106272, “Small- significantly as the acid spends, which Scale Fracture Conductivity Created by creates effective diversion of the acid Modern Acid-Fracture Fluids,” by M. in matrix acidizing. The same characPournik, SPE, C. Zou, SPE, C.M. Nieto, teristics have been found to reduce the M.G. Melendez, D. Zhu, and A.D. acid-leakoff rate and increase stimulaHill, SPE, Texas A&M University, and tion effectiveness in acid fracturing. X. Weng, Schlumberger, prepared The viscoelastic diverting acid system for the 2007 SPE Hydraulic Fracturing used in this study consists of 15% HCl, Technology Conference, College Station, 7.5% viscoelastic surfactant, and a corrosion inhibitor. Texas, 29–31 January.
Laboratory Scaling The acid-fracturing process involves several different phenomena, including acid transport along the fracture, acid transport into the rock because of leakoff, convective and diffusive mass transfer of acid to the fracture face, and acid reaction on the fracture face. The appropriate dimensionless groups that arise from conservation equations provide a guide for proper experimental scaling. For a typical field treatment with a 20-bbl/min injection rate into a fracture 100 ft high and 0.2 in. wide, the injection rate in the experiments should be approximately 2.2 L/min to create the same Reynolds number as would occur near the wellbore in the field fracture. As the acid moves down the fracture and leakoff occurs, the velocity and Reynolds number will decrease. An injection rate of 1 L/min was used in the experiments to ensure that the hydrodynamic effects that occur in the field were simulated in the laboratory. Fracture-Conductivity Experiments The experimental apparatus and procedure were designed to enable experiments to be conducted at conditions more closely and accurately representing field treatments than previous studies. The experimental apparatus also was designed to accommodate larger rock samples, higher injection rates, and higher temperatures. The core samples were placed inside a modified API conductivity cell with body dimensions of 10×31 / 4×8 in. and a 71 / 4×13 / 4-in. opening, allowing the use of rock samples with thicknesses as great as 3 in. Cylindrical ceramic radiant heaters that can heat the fluid to 300°F were wrapped around the flowline to heat the fluid before it entered the cell so experiments could be conducted at temperatures similar to field conditions. A backpressure regulator was installed on the leakoff line
For a limited time, the full-length paper is available free to SPE members at www.spe.org/jpt. The paper has not been peer reviewed. JPT • JULY 2007
to control pressure drop across the core. This allowed the leakoff flow rate to be controlled to represent field conditions more accurately. Another backpressure regulator was installed on the cell effluent line to maintain the cell pressure above 1,000 psi.
ments, the injection rate was 1 L/min with a 1,000-psi backpressure and leakoff flux of approximately 0.005 ft/min. The initial fracture gap was maintained at 0.12 in. The first three sets of experiments were conducted under the same conditions except that each set was performed with a different acid. These three sets constitute the basic data set for comparison and evaluation of acid type and contact time.
Experimental Procedure. An acid-fracture-conductivity experiment consists of core preparation, acid etching, rockembedment-strength measurement, surface-profile characterization, and con- Acid Type. For all contact times, the ductivity measurement. Core samples emulsified acid results in the lowest were cut into a parallelepiped shape conductivity at all closure stresses. The with the ends curved to fit the API cell. viscoelastic-diverting-acid and polyThe cores were approximately 7.11 in. mer-gelled-acid systems compete for long, 1.61 in. wide, and 3 in. thick. The highest created conductivity, dependcores were covered with a silicone-rub- ing on contact time and closure stress. ber compound to secure a tight fit of For a 15-minute contact time, the visthe core in the API cell and to prevent coelastic diverting acid yielded the highany acid bypass around the core sample. est conductivity at all closure stresses. The cores then were placed in a vacuum This can be explained from the etching device for several hours until all air was pattern, which shows a narrow-channel removed from the pore spaces, then the development with greater roughness cores were saturated with water. than the other acid systems. The other acid types dissolved less rock and thus Conductivity Measurement. Fracture created less conductivity. conductivity was measured by flowing However, at the 30- and 60-minute nitrogen between the two acid-etched contact times the polymer-gelled-acid core samples and recording the absolute system created a rougher etching pattern pressure at the midpoint of the fracture with a narrow channel, while the viscoand the pressure drop across the frac- elastic-diverting-acid system etched the ture. The conductivity cell was placed surface too much, resulting in a uniform in a load frame that provided closure etching pattern. The embedment strength stresses up to 6,000 psi. A 1,000-psi of the core acidized with viscoelastic closure stress was applied and increased diverting acid is almost half the strength in 1,000-psi increments to 6,000 psi, of the core acidized with polymer-gelled changing the load after approximate- acid. As a result, for a 30-minute contact, ly 60 minutes. Nitrogen flow was while the initial conductivity of the viscobegun after placing the first load on elastic-diverting-acid system at 1,000-psi the cell. Pressure readings were record- closure stress is higher than that created ed at four different flow rates. Flow by the polymer-gelled acid, at higher rates ranged from approximately 5 to closure stresses, the conductivity of the 20 L/min. Forchheimer’s equation for viscoelastic-diverting-acid system is less flow through a porous medium was used than the conductivity created by the to calculate the fracture conductivity. polymer-gelled-acid system. Similarly, for a 60-minute acid contact, the polyResults and Discussion mer-gelled acid creates higher conducFifteen experiments divided into five dif- tivity than the viscoelastic-diverting-acid ferent sets were conducted to study the system until the 5,000-psi closure stress effects of acid type, acid contact time, is applied. Above this closure stress, the temperature, and rock type on acid-frac- rock surfaces acidized by either system ture conductivity. For each of the three have been crushed, and essentially all acids, three different contact times were fracture conductivity is lost. used that ranged from 15 to 60 minutes. Temperature was maintained at 200°F Acid Contact Time. The change in for most experiments, except for one fracture conductivity with acid contact set with emulsified acid at 275°F. The time, and the interrelationship between experiments were conducted on Indiana contact time and closure stress, is perlimestone samples except for one set that haps the most interesting result of this used Silurian dolomite. For all experi- study. There appears to be a trade-off
between increasing initial (low-closurestress) conductivity as more rock is dissolved with longer contact times and weakening of the rock surfaces with longer contact times. The results for the viscoelastic-diverting-acid system gave the clearest indication of this effect. At a low closure stress of 1,000 psi, the conductivity increases gradually with increased acid contact time, ranging from approximately 17,000 md-ft for the 15-minute-contact-time experiment to approximately 36,000 md-ft for the 60-minute test. This behavior, conductivity increasing with increasing amounts of dissolution, is the behavior predicted by the Nierode-Kruk correlation. In contrast, at 3,000-psi closure stress, the conductivity declines from more than 12,000 to approximately 2,000 md-ft for this same range of contact times. The longer acid exposure time weakens the rock, and surface asperities are crushed at 3,000-psi closure stress. With 5,000-psi closure stress, only the samples that were acidized 15 minutes maintained any appreciable fracture conductivity. With 30 or 60 minutes of acid contact time, the rock surfaces apparently are so weakened that the fractures closed almost completely. The general trend of decreasing rock embedment strength with increasing acid contact time supports this interpretation. These results suggest that overtreating with an acid exposure time that is too long may yield lower fracture conductivity than can be obtained with shorter acid contact times, and that this effect is more important at higher closure stresses. This could be particularly harmful if the near-wellbore part of the fracture, which has the longest acid exposure time, has relatively low conductivity. However, it is possible that these effects observed in small-scale laboratory tests could be overcome by larger-scale channel features created at field scale. Temperature. Experiments with the emulsified-acid system were conducted at 200 and 275°F. Temperature had a profound effect on the conductivity created with this acid system; the conductivity at 275°F was approximately an order of magnitude higher than at 200°F for the entire range of closure stresses. Higher temperature resulted in much more dissolution and roughness on the fracture face compared with the 200°F experiments, while rock strength was JPT only slightly reduced.
JPT • JULY 2007
WE LL ST IMUL AT ION
Rock-Mechanics Considerations in Fracturing a Carbonate Formation Acid fracturing is used to improve well productivity in acid-soluble formations such as limestone, dolomite, and chalk. Proppant fracturing is an alternative option used in carbonate formations. There is no quantitative method to determine whether acid fracturing or proppant fracturing is an appropriate stimulation method for a given carbonate formation. Laboratory experiments were performed with full core samples to examine the effect of elastic, plastic, and viscoelastic rock behavior on fracture conductivity for acid- and proppant-fracturing treatments.
Introduction Hydraulic fracturing (acid or proppant) is used to create a conductive fracture in the formation to improve well productivity. The induced fracture tends to close because of the effect of the minimum horizontal stress. Fracture closure is controlled by elastic, plastic, and viscous rock properties. In acid fracturing, the etched, nonsmooth fracture surfaces leave open pathways upon closing in addition to the wormholes and channels created from the fracture into the formation. Fracture conductivity is generated by the quantity of rock removed and the rock-removal pattern. Depending on the pattern of the natural-fracture system, acid solubilThis article, written by Assistant Technology Editor Karen Bybee, contains highlights of paper SPE 102590, “Acid Fracturing or Proppant Fracturing in Carbonate Formation? A Rock Mechanic’s View,” by H.H. Abass, A.A. Al-Mulhem, M.S. Alqam, SPE, and K.R. Mirajuddin, SPE, Saudi Aramco, prepared for the 2006 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, 24– 27 September.
ity of the formation, magnitude of the minimum horizontal stress, and reservoir temperature, acid fracturing vs. proppant fracturing should be evaluated to select the most effective stimulation treatment for a given formation. Although longer acid contact with the formation results in more etched surface and thus higher fracture conductivity, formation compressive strength is reduced. Claims have been made that at high reservoir temperatures, fast acid reactions in formations containing high concentrations of calcite result in acid-fracture lengths much shorter than propped fractures. The suggestion was made that for reservoirs with a minimum horizontal stress (fracture-closure stress) greater than 5,000 psi, proppant fracturing is the optimum stimulation method because etching caused by fracture acidizing cannot support such high stress. In chalk formations, it has been shown that proppant fracturing yielded better results than acid fracturing. Fracture length from acid fracturing and proppant fracturing will be different because of their dissimilar fracture mechanics. In proppant fracturing, the fracturing gel is not reactive with the formation and can penetrate deeper when compared with acid fracturing for a given fracturing-fluid volume, especially at high reservoir temperature. Therefore, it is expected that proppant fracturing will create longer fractures. The full-length paper presents a rockmechanics view of fracture closure of propped and acid-etched fractures that describes the following. • In acid fracturing, fracture closure extent is the result of asperities embedment, asperities crushing, and viscous flow (creep).
• In proppant fracturing, fracture closure extent is the result of proppant embedment, proppant crushing, and proppant flowback.
Acid-Fracture Closure The increase in production from an acid-fracturing treatment is the result of fracture length and fracture conductivity. Fracture length is controlled by acid convection (injection rate), acid-reaction rate, and acid-loss rate. Fracture width is a result of the differential etching occurring as the acid reacts with the walls of the created fracture. This creates an uneven fracture surface that determines the fracture width upon fracture closure. Fracture conductivity is determined by the amount of rock dissolved, fracture-surface roughness, closure stress, and the stress/strain characteristics of the rock. If reservoir temperature is too high, injection-rate optimization becomes critical to create a long conductive fracture. If the reaction rate is low, uniform etching may result, leading to insufficient fracture conductivity. Upon completion of an acid-fracturing treatment, the following three factors contribute to a reduction in fracture conductivity. • Elastic response. • Compressive failure of contact points (asperities). • Creep effect. Elastic Response. The elastic closure response occurs when the net effective minimum horizontal stress increases as a result of reservoir depletion. The elastic response to close the fracture follows Hooke’s law of elasticity and is controlled by Young’s modulus of the formation. The elastic response will decrease the fracture aperture, which reduces fracture conductivity.
For a limited time, the full-length paper is available free to SPE members at www.spe.org/jpt. The paper has not been peer reviewed. 50
JPT • JULY 2007
Fig. 1—Fracture surface before acidizing (above) and after acidizing (below).
Compressive Failure. The compressive strength of the asperities will determine the severity of their failure on fracture permeability. The
JPT • JULY 2007
reduction in conductivity is a result of the combined effect of elastic response and compressive failure of the asperities. Compressive fail-
ure also generates rock particles and fines that reduce fracture conductivity further. Creep. The creep (viscous) effect is a slow, time-dependent displacement. The total displacement obtained from applying a constant stress is the sum of two components, displacement resulting from elastic response and the creep function. The creep function characterizes the rheological properties of the rock formation and is best described experimentally for a given stress range, temperature, and lithology. Creep models include the elastic response described by Hooke’s law for Hookean substances (spring model) and the viscous response for Newtonian substances as described by a dashpot model. Both of these effects act when the reservoir pressure decreases: the elastic displacement (spring effect) in response to the increase in effective closure stress, and a time-dependent displacement function (dashpot effect). All viscoelastic models include both effects to simulate a creep phenomenon.
horizontal fracture. The same design Experimental An experimental procedure was was used for a propped fracture; howdesigned to simulate acid- and prop- ever, a proppant layer was placed on pant-fracturing processes using a one surface. rock-mechanics loading frame. Two The other sample geometry was types of geometry were used that sim- created by splitting a 4-in. whole core ulate radial- and linear-flow regimes, into two halves longitudinally. Then, respectively. Whole core samples the confining pressure was applied were selected that were 4 in. in diam- around the fractured sample and lineter and approximately 4 in. long. ear flow was established to deterA 1 / 4-in.-diameter hole was drilled mine the conductivity of an etched or in the center so radial flow could be propped fracture. established through the rock matrix or through an induced fracture. Then, Creep Test. A creep test was designed the sample was cut horizontally into by applying in-situ conditions of two pieces to simulate a fracture. The temperature and stress for a given surfaces simulating a fracture were sample. Progressive loads simulating exposed statically to 15% acid by a stress path imposed on a fracture either dipping the sample in acid or during production were applied and placing acid on the surface until no maintained constant as the resulting more chemical reaction was observed. deformation was measured. Fracture Fig. 1 shows the fracture surface conductivity was calculated for the before and after acidizing. The sam- progressively applied stresses to ple was bound together again with determine its variation resulting from the same alignment before acidizing elastic, plastic, and viscous effects. by matching two lines drawn on the sample before cutting. Creep The final geometry of the experi- Acid-Fracture Sample Test. A creep ment was a vertical wellbore with a test was designed to study rock defor-
JPT • JULY 2007
mation under constant stress as a function of time. This test simulates the in-situ reservoir conditions where a fracture is exposed to the effective minimum horizontal stress. A typical test involves loading the sample to three progressive stresses: 4,000, 6,000, and 8,000 psi. At each stress level, the elastic and viscous displacements were measured until a trend was obtained. The creep profile obtained suggests that the sample exhibits primary and secondary creeping phases but has not shown any sign of tertiary creep. This is expected for such a high-Young’s-modulus sample. Creep Modeling. To model the complete creep response (primary and secondary), Burgers substance was used to describe the axial strain as a function of time for a sample sub jected to constant axial stress. This model includes the instantaneous strain, transient creep, and steadystate creep. The experimental creep data for 4,000-psi axial stress were matched by Burgers substance. The model clearly illustrates the nonlinear time-dependent behavior.
Fracture Width Fracture width varies significantly between acid fracturing and proppant fracturing. Fracture width in acid fracturing is created from the etching mechanism, and upon closing, channels are left open because of the nonsmooth surfaces of the created fracture. In proppant fracturing, a fracture closes on a proppant bed leaving a continuous highly permeable fracture (not channel) connecting the reservoir to a wellbore. The displacement resulting from creep compared to elastic response becomes significant with time. This displacement will not close the fracture directly, but it is manifested in stress applied on the contact points (asperities) in acid fracturing or on the proppant grains of the proppant pack in proppant fracturing. Conductivity Rock A. To evaluate the effect of elastic and creep displacements on fracture conductivity, flow testing was conducted with mineral oil. Production rate decreased from 180 cm 3 /min to approximately 20 cm3 /min as stress increased from 1,000 to 4,000 psi. The
JPT • JULY 2007
stress then was maintained at 4,000 psi to evaluate the creep effect over time. The production rate declined from 20 cm3 /min to approximately 5 cm3 /min after 100 hours. The creep effect can be less dramatic if the fracture can transfer the creep force through contact points without failure. As closure stress increases, some contact points fail and a continuous production-rate decline is anticipated. In a propped fracture, the creep force is transferred if the proppant-grain strength is sufficient, otherwise proppant crushing occurs. The effect of creep on proppant-fracture conductivity was not significant. For this formation, a proppant fracture will sustain well productivity while acid fracturing will suffer a production decline with time. Rock B. This sample was from a different formation. The effect of stress on permeability was evaluated for the matrix, tensile fracture, 100-mesh layer of 0.12-in. sand, one layer of 30-mesh resin-coated proppant (RCP), and an acid fracture. The permeability of one layer of 30-mesh RCP decreased drastically and extensive fines generation occurred at a 4,000-psi effective confining pressure. This is an important criterion to consider when deciding on the proppant type to be used in a proppant-fracturing treatment. In this formation, the acid fracture exhibited only a small decrease in permeability as a function of increasing stress.
Mechanical Strength of Fracture Surface Long acid-contact time may not be beneficial to fracture conductivity because it can weaken the fracture surface and make it more vulnerable to creep and contact-point compressive failure. It was shown that conductivity created by a 20-minute acid-contact time was higher than that created by 40-minute contact for both dolomite and limestone samples. The acid exposure weakened the rock structure along the fracture surface, resulting in greater sensitivity to closure stress. The fracture surface becomes more plastic, and the contact points will fail at higher closure stresses. This effect is more pronounced near the wellbore because that is where the acid-contact JPT time is the maximum.
WE LL ST IMUL AT ION
A New Method for Acid Stimulation Without Increasing Water Production Successful acid stimulation requires a method to distribute the acid between multiple hydrocarbon zones. Because almost all producing wells contain sections of varying permeability, this can be a problem. Because acid is an aqueous fluid, it tends to enter the zones with the highest water saturation. These water zones also are often the highest-permeability zones, so acid stimulation often will result in large increases in water production. The full-length paper describes use of a new low-viscosity system that reduces formation permeability to water with little effect on hydrocarbon permeability and also diverts acid from high-permeability zones to lower-permeability zones.
Introduction In matrix-acidizing treatments, the acid tends to enter the highest-permeability layers and bypass the mostdamaged (lower-permeability) layers. Various placement techniques have been used in attempts to achieve uniform placement of acid across all layers. The most reliable method uses mechanical isolation devices (such as straddle packers) that allow injection into individual zones until the entire interval is treated. However, this technique often is not practical, cost-effective, or feasible. Without a This article, written by Assistant Technology Editor Karen Bybee, contains highlights of paper SPE 103771, “A New Method for Acid Stimulation Without Increasing Water Production: Case Studies From Offshore Mexico,” by G.H. Reza, Pemex, and E. Soriano, SPE, L. Eoff, SPE, and D. Dalrymple, SPE, Halliburton, prepared for the 2006 SPE International Oil Conference and Exhibition, Cancun, Mexico, 31 August–2 September.
packer, some type of diverting agent must be used. Typical diverting agents include ball sealers, degradable particulates, viscous fluids, and foams. Although these agents have been used successfully, all have potential disadvantages, and none address the prob-
lem of increased water production that often follows acid treatments. One method of controlling water production uses dilute polymer solutions to decrease the effective permeability to water more than to oil. These treatments are referred
For a limited time, the full-length paper is available free to SPE members at www.spe.org/jpt. The paper has not been peer reviewed. 54
JPT • JULY 2007
to as relative permeability modifiers (RPMs), disproportionate-permeability modifiers, or bullhead treatments. RPM systems are thought to perform by adsorption onto the pore walls of the formation flow paths. Previous papers have described the development of and laboratory studies of an RPM based on a hydrophobically modified, water-soluble polymer, referred to as an associative polymer (AP). This group of polymers was selected for study because their properties can be altered in ways that render them valuable for oilfield applications. Another paper has described a laboratory study of this polymer for use as an acid diverter.
polymer onto a surface. In general, hydrophobic modification of watersoluble polymers adds new properties while retaining features typical for hydrophilic polymers. Viscosified or foamed fluids commonly used for acid diversion can result in high frictional pressure loss and require special manifolding and/ or pumping equipment. The low viscosity of the AP diverting system results in ease of mixing, low-frictional pressure losses, and no special manifolding or pump requirements. The diversion of aqueous fluids occurs only after the material enters the porous media, whether it is naturally fractured carbonate/dolomitic rock or sandstone matrix. It is theorized that the increased shear encountered upon entering the rock matrix, coupled with polymer adsorption, results in an apparent “viscosity” increase that may be responsible for the pressure increases seen during the treatment.
AP Properties The solution properties of both ionic and nonionic, water-soluble polymers are uniquely modified when hydrophobic groups are introduced into the polymer chains. The primary factor responsible for the property modification is the associative Job Results tendency between the hydrophobic More than 30 wells have been acidized groups when placed in an aqueous with the AP diversion system in the medium. Previous testing has shown Chuc, Caan, and Pol fields. These a unique shear-thickening phenom- fields are primarily dolimitic, and enon for the AP used in the current the acid has consisted of hydrochlowork. However, the solutions used ric/acetic blends formulated specifiin diversion operations show very cally to avoid sludging problems. low viscosity (less than 2 cp) at Results from nine of these jobs are surface conditions. shown in Table 1 in the full-length The adsorption behavior of hydro- paper. The oil- and water-producphilic water-soluble polymers also tion numbers shown are the approxican be modified in a unique manner mate values just before and after by the introduction of hydrophobic acid stimulation. groups. Rather than reaching plateau adsorption, as is common for Without Diverter. Two acid jobs hydrophilic polymers, hydrophobic were performed on Well Caan 53. modification appears to produce a Following the first job in August continued growth in adsorption with 1999, the water cut began to increase, increased polymer concentration. followed by a decline in the oil rate. This behavior is attributed to associa- A second acid job was performed tive adsorption of polymer chains on on a different interval in December previously adsorbed layers of poly- 2003, again followed by an eventual mers. Fig. 1 illustrates the adsorption increase in the water cut. So, while of a nonmodified and a modified the acid jobs performed on this well
JPT • JULY 2007
increased oil production, increases in 800 BOPD with 50% water cut, and water production also were seen. these levels were holding steady more On Well Caan 51, the interval from than 2 years later. 12,687 to 12,737 ft showed an increase In Well Chuc 173, the original perin water cut in March 2001, with a con- forated interval began to show an oil current decrease in oil production. An production decline in mid-2002. In acid-stimulation job in September 2003 November 2005, the interval from did not increase the oil rate, although 14,020 to 14,108 ft was perforated and the water rate continued to increase. acidized. While there was an immediThis interval was isolated, and the ate increase in oil production, there interval from 12,255 to 12,333 ft was also was an onset of water production. perforated and acid stimulated. This These results are typical of acid-stimuresulted in an increase in oil, although lation jobs in the Caan, Pol, and Chuc approximately 1 year later the water cut fields. Other diverters were used in this also began to increase rapidly. field, such as ball sealers and foam, On Caan 96, an acid-stimulation but with no real benefit. Also, because treatment was performed in March of the close proximity of water zones 2003 and resulted in no increase in in many wells, the acid volumes were oil production. In December 2005, the reduced in an attempt to avoid the water cut on this well began a sharp onset of water production. increase. On Well Pol 388, the interval from With Diverter. In Well Caan 73A, 13,715 to 13,796 ft was producing the interval from 13,190 to 13,222 ft 2,800 BOPD at approximately 50% showed a sharp increase in water prowater cut in August 1992. The well was duction in January of 2003, along with shut in until November 2003, at which a sharp decrease in oil production. This time the interval from 13,715 to 13,796 interval was isolated, and the interval ft was perforated and acid stimulated. from 13,098 to 13,131 ft was perforated Initial production was appproximately and acidized with the AP diverter. Even
with the close proximity of a water-producing zone, this interval has produced water-free for more than 2 years. In the Chuc 63 well, the initial production was approximately 1,000 BOPD, with a 16% water cut. After 1 month, production was approximately 1,800 BOPD with the same water cut. However, less than 1 month later, oil production dropped dramatically. The well was acid stimulated in February 2005 with the AP diverter. The increase in oil production was excellent. One year later, the oil production appeared to be remaining steady at almost 3,600 BOPD. In addition, rather than increasing the water production, the water cut fell from the initial level of 16% to approximately 4% and also was holding steady 1 year later. In Chuc 192, oil production began to drop rapidly with a concurrent increase in water production in February 2005. This interval was isolated and a new interval perforated and acid stimulated with the AP diverter. Oil production increased along with an approximately 2% water cut, which fell to zero within a few months. The increased oil production with no water has held steady for almost 1 year. Out of nine wells acidized with the AP diverter, the average increase in oil production was 231%. In addition, 22% showed a decrease in the water cut following the job, and 67% showed no change in the water cut following the job. For the five wells acidized without the AP diverter, four showed substantial increases in water cut following the jobs. In fact, one well went from producing no water to producing 3,273 BWPD. Also, three of the wells showed decreased oil production following the jobs.
Conclusions 1. Laboratory tests have shown that the AP diverter can divert acid from predominantly water-saturated zones to predominantly oil-saturated zones in both sandstone and carbonate lithology. 2. In sandstone and carbonate, the AP diverter can provide acid diversion and permanent water-permeability reduction. 3. Results from the Chuc, Caan, and Pol fields show that use of the AP diverter results in lower water production and increased oil production compared with control wells acidized JPT without the AP diverter.
JPT • JULY 2007
WE LL ST IMUL AT ION
Horizontal-Openhole Gravel-Packing Operations in the Campos Basin The full-length paper presents an overview of the evolution of openhole gravel-packing practices and experience after 200 wells have been completed successfully with this technique in the Campos basin. A comprehensive description of the main steps taken to improve horizontalopenhole gravel-packing (HOHGP) practices in unconsolidated oil-bearing turbidites is presented. Since the first HOHGP job in 1988, completions have moved progressively from shallow- to ultradeepwater scenarios. Along this path, a series of innovations has been incorporated into the sandface-completion practices.
Introduction The most prolific reservoirs in the Campos basin are the Upper Cretaceous and Tertiary turbidites. The high-permeability (approximately 1,000 to 8,000 md) stacked and amalgamated reservoirs are spread over shallow, deep, and ultradeep water within the basin. Dictated by the depositional model associated with turbidites, the sand uniformity of these poorly- or unconsolidated sand lenses varies significantly. The presence of reactive-shale streaks is recurrent in some of these turbiThis article, written by Assistant Technology Editor Karen Bybee, contains highlights of paper SPE 106364, “The 200th Horizontal-Openhole Gravel-Packing Operation in Campos Basin: A Milestone in the History of Petrobras Completion Practices in Ultradeep Waters,” by L.C.C. Marques, SPE, L.C.A. Paixão, V.P. Barbosa, M.O. Martins, A. Calderon, SPE, C.A. Pedroso, SPE, L.H.C. Fernandes, J.A. Melo, C.M. Chagas, and N.J. Denadai, Petrobras, prepared for the 2007 SPE European Formation Damage Conference, Scheveningen, The Netherlands, 30 May–1 June.
dites. As in many other offshore basins in the world, the first oil discoveries (early in the 1980s) were in shallow waters. These good exploratory results propelled the move progressively from shallow to ultradeep water. However, since the original oil discoveries, it has been realized that a sand-management strategy was necessary to achieve desirable production levels. Sand control is an umbrella term comprising different approaches to dealing with sand-production problems. Sand-control methods include frac pack, chemical consolidation, screens, and gravel packing. The Petrobras philosophy is one of zero tolerance for sand production. Should there be the slightest chance of sand production, a sand-control method is installed in wells. This preventative approach stems from wellbore-integrity concerns, prohibitively high wellintervention costs, the need to maximize production rates, safety concerns, and the inability of topside equipment to handle sand. Gravel packing is considered the best alternative for sand control in horizontal-openhole wells with good vertical permeability, nonuniform sands, and no lamination. In addition, filling the screen/wellbore annulus of an openhole horizontal well with properly sized gravel creates a secondary barrier to the migrating sand grains, thus increasing the longevity of the gravelpack screens.
Sand-Control History The ability to achieve reliable sand control and high completion-efficiency indices is the result of a philosophy of teamwork that integrates drilling and completion practices. Use of sand-control advances provided by the sand-control industry has enabled Petrobras to set the pace in sand-con-
trol and completion practices in deep and ultradeep water. These include use of premium screens and tailor-made drill-in fluids, application of absolute filtration standards for completion and gravel-pack fluids, use of improved wellbore-cleanup procedures, extensive use of core-flow studies to minimize formation damage by drill-in fluid, optimization of bridging-material size distribution, use of drillpipe with internal plastic coating to prevent rust from being carried to the formation face, achieving a better understanding of the rock mechanics necessary for horizontal-openhole-well construction, use of stainless-steel gravel-packing pumping equipment, use of low-density gravel, and use of modern geosteerable tools to reduce wellbore tortuosity and to drill in-gauge wells. In the early 1980s, only shallowwater cased wells were completed with the gravel-packing technique. The achieved completion-efficiency indices were quite low and posed a major threat to the economic feasibility of ongoing projects in deep waters. Because of this, a sensitivity analysis was run to bracket the effect of different variables on the production impairment observed. Frac Packs.
The frac-pack technique was the first sand-control option used to complete vertical and low-angle deviated cased wells with no mechanical restriction and with no gas/oil or water/oil contacts nearby. As predicted by a suite of studies, use of the fracpack technique in the Marlim field has produced completion-efficiency indices as great as 100% in several situations as the result of creating a highly conductive fracture in the formation that balances flow restrictions that occur in the perforations. However, HOHGP wells have been adopted as the best-in-class well architecture for
For a limited time, the full-length paper is available free to SPE members at www.spe.org/jpt. The paper has not been peer reviewed. JPT • JULY 2007
the newest sand-control projects in the Campos basin. Because of this, frac packs have been limited to some very specific cased-well applications. Expandable Sand Screens (ESSs). Since the early 2000s, ESSs have been installed in horizontal-openhole injectiors and producers in the Campos basin. Unexpectedly poor results were obtained from this initial experience. Premature sand failure occurred in most of the producers completed with ESSs. ESS longevity in producers did not meet expectations.
Gravel-Packing Simulators Although gravel placement in horizontal wells is a mature technology, the commercially available numerical gravel-packing simulators are empirical and not fully validated for ultradeepwater applications where low fracture gradients dictate a narrow operational window. Neither have they been validated for lightweight gravel materials nor for Newtonian solids-free synthetic fluids used as gravel-carrier fluids. Improvements are still needed in pressure-behavior analytical tools to simulate gravel-packing operations. If the appropriate simulation capabilities were available, they would enable real-time decision making and running post-job analyses to match the predicted pressure-behavior profile with actual field measurements. A faster solution could be to upgrade existing gravelpacking simulators in such a way as to accomplish this. Occasionally, downhole-pressure gauges have been run in the hole during gravel-packing operations to gather data to calibrate and tune the gravel-packing simulator. Drill-In Practice Improvements One the keys to achieve damage-free HOHGP is the correct choice of the drillin-fluids program. A comprehensive well-engineered drill-in-fluids program must consider all the steps between the moment the drill bit tags the pay zone and the moment at which the solidsladen drill-in fluid is replaced by a solids-free fluid before the gravel-packing operation. An in-house research program was created to investigate the influence of different variables on drillin-fluids performance. The stability of a horizontal-openhole well is dependent on maintaining an adequate hydrostatic head on
the formation, creating a resistant mudcake on the formation face being drilled, and inhibiting shales. In some shallow-water situations, the mudcake must be able to resist a very high differential pressure (more than 1,000 psi from wellbore to formation) and still have good leakoff control to prevent deep formation invasion by the particles. This is particularly important because of the high permeability of the Campos basin turbidites. In addition, the mudcake particles must be sized to flow through the gravel pack, across the screens, and into the wellbore when the well is put on stream.
Gravel-Carrier-Fluids Filtration Improvements The importance of clean fluids to prevent in-flow formation damage is well established. Since the early 1990s, absolute filtration guidelines were established for gravel-carrier fluids. A minimum beta ratio of 5,000 (at 2 µm) is the filtration-efficiency level stipulated for these fluids. Typically, some thousand barrels of filtered brine are spent in the drill-in fluids replacement, wellbore cleanup, and gravel-packing operations. The offshore logistics of preparing such a large volume of fluid poses enormous practical problems. Small rig-pit capacities for ultradeep water are common. Therefore, it is more feasible to produce these fluids [saturated sodium chloride (NaCl) brine] in an offshore facility (fluid processing plant) and ship them just in time for the operations. Once on the rig, the brine then is absolute filtered, blended with chemicals, and eventually diluted with filtered seawater. Good transportation logistics is necessary to prevent fluid contamination and delays. Drill-In-Fluid Improvement Tailored Drill-In Fluids. The use of water-based fluids has been a rule in these horizontal-openhole drill-in operations. These tailor-made fluids are composed of an NaCl brine, a shearthinning biopolymer, a lubricant, a loss-control additive, a biocide, an alkalinizer, a temperature-activated enzyme breaker, and HCl-soluble aragonite particles sized to prevent deep invasion of the formation pores. A current drilling practice in the Campos basin is to drill the 12 1 / 4-in. phase with a synthetic oil-based mud
(SOBM) to target. A pilot well is first drilled to obtain geological data on the sand layers, thus helping to define the horizontal-well trajectory. Then, the 95 / 8-in. casing shoe is set and cemented at the top of the uppermost layer of the turbidites. Once the 121 / 4-in. drilling phase is finished, all the SOBM is replaced by a water-based drill-in fluid. The major drawbacks of such a fluid-replacement practice are potential formation damage, complex logistics, and time required. SOBM for Drill-In Fluids. Using SOBM to drill the 81 / 2-in. phase to construct a horizontal-openhole well to be gravel packed presents the following advantages over the current field practice of replacing this fluid with a water-based drill-in fluid: superior wellbore stability, better lubricity, higher rate of penetration, better shale inhibition, and rig-time optimization. Laboratory experiments and full-scale simulations were performed to specify the SOBM formulation to drill the 81 / 2-in. phase.
Bottomhole-Pressure Reduction Lightweight Gravel. Commercially available high-permeability synthetic proppants [16/20- and 20/40-mesh, specific gravity (SG) from 2.65 to 2.73] have a good record as a gravel material in conventional HOHGP operations. Where concerns exist in terms of conventional gravel placement—the existence of a washed-out zone, a low fracture gradient, long open hole to be completed, and fluid-returns problems—use of lightweight proppant (1.25 SG) as a gravel material has widened the operational window of HOHGP operations. Fluid Returns. Another solution to the low-fracture-gradient problem was use of a new flow path for gravel-carrier-fluid returns that resulted in a 500-psi pumping-pressure reduction at 8 bbl/min. Fluid returns to the surface were routed into the flowline instead of into the small-diameter chokeline and kill line. Once at the surface, the return fluid is not diverted to the trip tank to measure the flow rate but through a flowmeter installed in the flowline just before the gravel-packing operation. Flowmeters that are not affected by sea heave were designed to operate on JPT these floating rigs.