Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Upstream
Summary. Shutting off high-pressure water flows and achieving good cement bonding during primary cementing has long been a problem experienced by operators. Corrosion, casing leaks, excessive water production, and contamination of freshwater aquifers have led to costly workover operations throughout the oil field, especially within the Permian Basin. Remedial cementing jobs are often expensive and unsuccessful when primary cementing techniques fail. Therefore, a cost-effective, engineered method for cementing off these water flows was developed. This paper describes the cementing techniques, well-control criteria, well preparation, and data evaluation. Use of new and existing cementing practices provided the flexibility necessary to handle variable well conditions. Preplanning, evaluation of produced waters, compatibility studies, and proper slurry design helped to eliminate poor cement jobs in many areas. If one works closely with service and rig personnel, the procedure can be accomplished quite effectively, even on wells in which unexpected water flows are encountered. Examples of the procedure are documented with proven field results under variable conditions. Introduction The need to shut off water flows during primary cementing is critical. Most regulatory agencies currently require protection of all freshwater zones and prohibit pressure on casing-string annuli. Casing corrosion, out-of-zone crossflow, mud contamination, and excessive hole washout are often the result of improperly cemented water flows. In most cases, improperly cemented water flows do not stop the actual flow and continue to act dynamically by channeling to other zones. Continual dissolution of salt zones and annular casing pressure typically result, often leading to casing corrosion and failure. These problems can increase drilling costs significantly and often require remedial cementing procedures. Experience in several areas has shown that remedial cementing work is expensive and time-consuming, especially when washed-out salt sections and corroded casing are involved. Existing methods for controlling water flows have met with limited success. The need to develop a simple, cost-effective procedure to handle a multitude of complications that can occur during cementing-including lost circulation, water-flow/cement incompatibility, and associated gas-was needed because these complications make most conventional cementing techniques impractical. Therefore, we decided to develop a cementing procedure that was similar to existing well-control methods to handle these procedure that was similar to existing well-control methods to handle these complications. Use of constant-bottomhole-pressure (BHP') well-control methods and sound cementing practices led to the development of a technique to control water flows during cementing. Review of Existing Procedures Over the years, several methods for controlling and cementing off water flows have been used. All have been used frequently, but their applications are limited and their success rates are poor. The most commonly used technique is the barrel-in/barrel-out method, which holds backpressure on the annulus during cementing. Water influx is limited by holding pressure on the annulus based on volume. Pressure is held with a choke so that the annular returns are equal to those of the cement and displacement mud pumped. The main problem with this technique is that it does not allow for gas expansion. In nearly all the cases studied, gas was present (mostly in small quantities) in the water flows. This can result present (mostly in small quantities) in the water flows. This can result in excessive backpressure and loss of circulation. In addition, if returns are lost while cementing with this procedure, the volume control is lost and no sound way of minimizing the water influx exists. Another common method of cementing water flows is to cement the well conventionally and then bullhead cement from surface down the annulus. This procedure usually results in good cement bonding above and below the water flow but none across it. Corrosion in the uncemented section is then accelerated and remedial cementing is often required. The use of external casing packers (ECP's) and multiple-stage cementing tools [more commonly called diverting valve (DV) tools] can minimize the uncemented area but do not actually shut off the water flow. Hence, the water flow results in the fracturing and flooding of the nearest permeable interval if the pressure is sufficient to fracture the zone. This is rarely a desirable alternative. Other problems associated with this procedure include cement contamination and formation breakdown above the procedure include cement contamination and formation breakdown above the water flow while bullheading. In several areas, regulatory agencies prohibit bullheading practices because of these problems. Other methods of prohibit bullheading practices because of these problems. Other methods of controlling water flows include openhole cementing and polymer squeeze jobs. Both have been used successfully but are often expensive and time-consuming. Also, success is very difficult to achieve in long washed-out intervals. As can be seen, all of these procedures have applications, but wellbore conditions are rarely ideal for their use. Hence, a cementing procedure to handle a multitude of wellbore conditions was required. Fracture-Gradients/Constant-BHP Procedures The planning process for wells in water-flow areas should include defining possible water-flow zones and their associated surface pressure and BHP. possible water-flow zones and their associated surface pressure and BHP. It also should include fracture gradients for zones that will be exposed before cementing. If this information is not available, we recommend that pressure-integrity tests be performed to +/-200 psi above the anticipated pressure-integrity tests be performed to +/-200 psi above the anticipated water flow surface pressure. Tests should be run at the previous casing shoe and at suspected weak zones. This procedure is not recommended for areas in which the formations do not heal after a leakoff is obtained. Determination of fracture gradients before the flow is encountered is very important because it will help to predict crossflow zones and to determine whether a two-stage cement job is required, The well should also be shut in after the flow is encountered to determine the actual pressures if well conditions are conducive to this practice. Many water flows are a direct result of water injection out of the intended zone in nearby wells. We recommend that injection wells in the vicinity of the subject well be shut down. This has helped to reduce the water-flow pressures and rates in some cases. We also advise a water-flow sample for compatibility testing. The water may pick up soluble minerals that can be incompatible with a particular cement system. If the water sample contains any chloride or magnesium, it should be tested in 5 to 25 % volume concentrations with the selected slurry because this concentration will most likely accelerate the setting process. Constant-BHP Procedures. Two constant-BHP methods were used to shut off water flows successfully during primary cementing operations. They are both simple and consistent with accepted well-control procedures. Each procedure requires some planning, but both are adaptable to several complications, such as loss of circulation and associated gas. SPEDE P. 191
- North America > United States > Texas > Permian Basin > Yeso Formation (0.99)
- North America > United States > Texas > Permian Basin > Yates Formation (0.99)
- North America > United States > Texas > Permian Basin > Wolfcamp Formation (0.99)
- (22 more...)
- Well Drilling > Pressure Management > Well control (1.00)
- Well Drilling > Casing and Cementing (1.00)
Summary Permeability estimates from the pressure derivative or the slope of thesemilog plot usually are considered to be averages of some large ill-definedreservoir volume. This paper presents results of a study of the averagingprocess, including identification of the region of the reservoir thatinfluences permeability estimates, and a specification of the relativepermeability estimates, and a specification of the relative contribution of thepermeability of various regions to the estimate of average permeability. Thediffusion equation for the pressure response of a well situated in an infinitereservoir where permeability is an arbitrary function of position was solvedfor permeability is an arbitrary function of position was solved for the caseof small variations from a mean value. Permeability estimates from the slope ofthe plot of pressure vs. the logarithm of drawdown time are shown to beweighted averages of the permeabilities within an inner and outer radius ofinvestigation. permeabilities within an inner and outer radius ofinvestigation. The estimate is shown to be influenced most strongly bypermeabilities at a distance r=0.015 ()1/2 [5.6 ร 10โ5]. permeabilities at adistance r=0.015 ()1/2 [5.6 ร 10โ5]. Fully 50% of the contribution to thepermeability estimate is from the reservoir volume in the range from r = 0.011()1/2 [4.1 ร 10โ5] to r=0.022 ()1/2 [8.1 X 10โ5]. An analytic expression forthe weighting function is derived in this paper. Introduction Clues to the volume over which permeability averaging occurs can be obtainedfrom radius-of-investigation estimates and from studies of compositereservoirs. Certainly, regions in which fluids do not flow cannot contribute topermeability estimates. Thus, the drainage radius provides an outer bound tothe region of the reservoir that can influence the permeability estimate. Anestimate of the inner boundary of the region of influence can be obtained fromstudies of pressure-transient testing in reservoirs with a circulardiscontinuity in permeability. If porosity, compressibility, and thickness areconstant, then the slope of the first semilog straight line gives the mobilityof the inner cylindrical region that surrounds the wellbore, and the slope ofthe second semilog straight line gives the permeability of the outer region. Afairly long transition period connects one straight line to the other, even forabrupt mobility changes. Because the slope of the second straight line isindependent of inner-region permeability, we can conclude that the slope is notreflecting any simple average of all permeabilities inside the drainage radius. The question, then, is this: what averaging or smoothing process causes anabrupt property change to result in a smoothly varying slope on plots ofpressure variation? And for the more general case of continuously varyingpermeability, what does the slope of the semilog plot tell us about thepermeability distribution? Unfortunately, this averaging concept is somewhat incompatible with thecommon assumption that permeability is constant over large volumes of thereservoir. It is only when we consider that permeability is not constant thatthe averaging process becomes important. Little has been done to solve theproblem of pressure drawdown in a reservoir characterized by continuouslyvariable permeability or mobility. Exceptions include solutions to the problemof diffusion in a medium where permeability obeys a powerlaw dependence ondistance from the wellbore and Hantush's solution for the problem of unsteadyflow to a well in a reservoir where thickness varies exponentially in onedirection. In certain cases (e.g., the flow of gas in porous media), thevariable mobility is predominantly a function of pressure, and the nonlinearitycan be predominantly a function of pressure, and the nonlinearity can beremoved through the use of pseudopressures and pseudotimes. Other cases, including water coning, cannot be handled so simply. Despite the relaxation ofthe assumption of constant properties, applications of these methods are stillcharacterized by a very few degrees of freedom. To allow for the most generalpossible permeability distribution, a general solution to the diffusionpermeability distribution, a general solution to the diffusion equation isrequired. Although the problem of estimating an arbitrary permeability distributionfrom pressure data has not received much attention in the well-testingliterature, it has been actively investigated in the context of reservoirsimulation. In most cases, however, consideration is given to problems with along (i.e., years) history and a large number of wells. In addition, moststudies have focused on mathematical techniques for estimating the permeabilitydistribution and not on the problems of determining uniqueness or ofinterpreting the resulting model. Exceptions include a study by Jahns in whichthe uniqueness of permeability models determined from interference data werediscussed. Reciprocity, incomplete data, and parameter correlations wereidentified as factors that contribute to the nonuniqueness of parameterestimates. Jacquard and Jain pointed out that a proper geometric subdivision ofthe unknown permeability zones is not given proper geometric subdivision of theunknown permeability zones is not given a priori and that the uncertainty inpermeability estimates is a function of distance from the nearest well. Dogruet al. investigated confidence limits on parameter estimates resulting fromhistory matching the pressure data in single-phase reservoirs. Dogru andSeinfeld obtained an approximate solution to the problem of arbitrary radialand vertical permeability distribution by discretizing the spatial domain andpermeability distribution by discretizing the spatial domain and approximatingspatial derivatives with finite differences. They used the solution tocalculate sensitivity coefficients for the optimization of well-testdesign.
- Energy > Oil & Gas > Upstream (1.00)
- Government > Regional Government > Asia Government > Middle East Government > Saudi Arabia Government (0.47)
- Asia > Middle East > Saudi Arabia > Saudi Arabia - Kuwait Neutral Zone ("Partitioned Zone") > Arabian Gulf > Arabian Basin > Arabian Gulf Basin > Safaniya Field (0.99)
- Asia > Middle East > Kuwait > Saudi Arabia - Kuwait Neutral Zone ("Partitioned Zone") > Arabian Gulf > Arabian Basin > Arabian Gulf Basin > Safaniya Field (0.99)
Summary. Rheological data from cementing-spacer fluids are usually based on laboratory-prepared samples of the spacer and used to determine the flow rate required to provide the flow regime necessary for maximum efficiency of the spacer and to estimate the annular friction pressures associated with the spacer. Actual rheological properties are rarely measured on location, and the problems experienced with settling, foaming, and mixing are difficult problems experienced with settling, foaming, and mixing are difficult to simulate in the laboratory. This paper reviews the field data from three commercial spacers mixed at offshore locations and compares them with laboratory data for base spacer materials and weighted spacer mixes. The flow rates required to obtain turbulent and plug flow for each spacer are compared, and settling, foaming, and mixing problems are discussed. Because of solids settling in either the slugging pits or surface samples, the spacer formula is sometimes modified on location to improve solids suspension. The rheological effects of increasing concentrations of base spacer material, bentonite, and polyanionic cellulose (PAC) are discussed with a comparison of the rheological data from these spacers. Introduction Some scientists 1โ7 have indicated that maximum mud removal can be obtained through the use of a turbulent-flow spacer. Plug-flow systems have also been used successfully in many field applications. Other scientists have stated that regardless of the flow regime, the best mud removal is obtained by maximizing the flow rate in the well. Turbulent-flow-system design depends on spacer systems that possess sufficient suspension properties to prevent solids settling possess sufficient suspension properties to prevent solids settling on surface, but that either go easily into turbulent flow in the annulus, or at least minimize the friction pressure in the annulus at high pump rates. Because of the critical balance of these parameters, the quality of the base spacer material and of field parameters, the quality of the base spacer material and of field mixing of the spacer must be maintained at a high level. The surface rheology of a plug-flow spacer should be sufficient to suspend the weighting agent, but the ability to retain this viscosity downhole and to minimize annular friction remains a concern. This paper outlines the factors that can lead to variations in spacer rheology-particularly mixing equipment, personnel, and materials quality-and discusses the performance characteristics of the spacer on the cement job. Mixing Equipment More control can be kept over the type of mixing equipment used in spacer mixing for land-based than for offshore operations. The service company will generally provide the pumping equipment, which is normally the same equipment used to mix the spacer and the cement. Also, the spacer dry mix and weighting agent are normally blended at the service company bulk plant and delivered to the location as a complete mix. Offshore rigs lack sufficient space for additional mixing equipment and the required bulk capacity for holding a preblended spacer mixture, so sacks of dry spacer material are delivered with the appropriate mixing instructions for batch mixing in the rig's slugging pit. Barite is supplied by the rig for weighting the spacer to the desired density. These factors limit the amount of quality control possible at offshore locations. Also, mixing equipment and the amounts of materials added to make the spacer are not standard. The lack of standardized mixing equipment causes wide variations in the amount of shear applied to the spacer and in the final spacer rheology from one rig to another. Depending on spacer composition, the sensitivity to the amount of shear and the length of time shear is applied can drastically affect spacer properties. The problem with slugging pits usually is not excessive shear but lack of sufficient shear to yield spacer quickly. The spacer is therefore subjected to a longer low-shear period than in the laboratory. Because spacers are typically mixed m the slugging pit, the cleanliness of the mixing equipment may be questionable. In some situations, removing all the drilling mud from the mixing system can be difficult, and some residual mud may be found in the mixing system when the pumps are turned on to mix the spacer. Unless the amount of mud is excessive, slight contamination should not significantly affect the spacer. The initial rheology, however, can be affected, depending on the degree of contamination. Personnel Personnel The personnel charged with mixing the spacer offshore are usually not associated with the cement service company. Often the mud engineer and/or a member of the rig crew is responsible for spacer mixing but may not be familiar with the spacer or its expected performance in the well. In many instances, the only information the performance in the well. In many instances, the only information the mud engineer has is the amounts of water and spacer to mix. From there, the mud engineer must mix the spacer to the proper initial viscosity, increase the weight to the planned density without any solids settling, and pump the spacer into the well. The cement service company personnel are rarely involved in spacer mixing. They often have limited knowledge of the rig mixing equipment and are required to prepare for the cement job. Because of the differences in mixing equipment used offshore and because untrained personnel usually mix the spacer, the final spacer mixture and its properties are not consistent under field conditions. When the final spacer rheology is critical, additional care must be taken to ensure that the spacer has the properties specified in the plan, which requires that the personnel mixing the spacer know the plan and the actions required to obtain the desired results. Sauer recommended that a member of the cement service company be responsible for mixing the spacers, that spacer rheology be measured after mixing, and that the critical flow rates be recalculated on location. During this project, some spacers were mixed with the guidance of the service company representative, but the results still varied. The differences in final spacer properties appear to result from the inconsistencies of materials and mixing equipment rather personnel. Without the proper controls over spacer properties and a consistent method for obtaining these results, properties and a consistent method for obtaining these results, spacer properties will vary regardless of who mixes the spacer. Materials Quality Control Although quality control of the base spacer material should be checked by the service company, storage on the rig can alter its quality. Exposure to moisture can be detrimental to the spacer quality, and care must be taken to provide either dry, or at least covered, storage areas. Limiting the amount of spacer material in inventory can solve storage problems and can ensure fresh material for each job. Differences in spacer quality can also be related to the mix water available on location. Some polymers used in spacers can be sensitive to pH, salinity, and specific ion concentrations. Some either will not fully yield or can actually crosslink if the water quality is too far from specifications. The volume of water used to mix the spacer is equally important. Slugging pits are not usually calibrated to the accuracy of cement-batch-mixing equipment; thus, the final volume of mix water can vary. Some spacers appear to be much more sensitive to water quantity than others. SPEDE P. 196
Summary. This paper describes experiments with a large-scale flow loop (45 ft of 7-in. pipe in nominal 8 1/2-in. open hole) that measure the velocity of the interface between fluids of varying densities and rheologies. The work showed that single-fluid velocity profiles cannot be used to predict interfacial profiles (e.g., between spacer and cement) and confirmed the importance of turbulent flow in minimizing channeling through mechanisms of increased frictional pressure drop a nd mixing across the annulus. Large density differences, both negative and positive, also minimized channeling. Introduction Successful primary cementing is as important today as it has been in the past. A large percentage of wells are being drilled in the difficult offshore environment, where the cost of remedial work is significant, leading to renewed efforts to ensure efficient primary-cementing. Placement is a key part of the primary-cementing primary-cementing. Placement is a key part of the primary-cementing process and has attracted much attention in the past. As a result of process and has attracted much attention in the past. As a result of significant improvements in our understanding, a greater percentage of wells are now cemented efficiently. Nevertheless, a significant number of primary cement jobs still experience problems, particularly in such primary cement jobs still experience problems, particularly in such difficult situations as highly deviated wells or wells prone to severe washouts. The cost of failure in these wells is often extremely high. A quantitative understanding of the placement process, including the ability to predict performance, is required to achieve successful cement jobs over the wide range of conditions currently found in the field. BP Exploration built a large-scale wellbore simulator to improve the qualification of cementing variables. Current practice relies on a number of well-worn guidelines that are not universally applicable. When these guidelines are applied to some of the difficult cases, other important variables can be forgotten, resulting in poor cement placement. Lockyear and Hibber described the wellbore simulator and some preliminary results. This paper addresses the problem of channeling, particularly attempts to quantify the behavior of the cement/spacer and spacer/mud interfaces. Previous Studies Previous Studies Probably the most widely known guideline to achieve good cement placement Probably the most widely known guideline to achieve good cement placement is to pump in turbulent flow. This was first recommended by Howard and Clark, who displaced mud directly by cement without use of a spacer. Current practice is to use a spacer fluid to separate mud and cement, but when a spacer is used, it is often unclear which fluid needs to be pumped in turbulent flow. In contrast, Parker et al. suggested that the viscous reaction between mud and cement could be used to obtain good cement placement if the overall velocity were less than 90 ft/min. This idea led to the plug-flow technique and the recommendation that under such circumstances the cement needed to have a specific gravity (SG) 0.24 greater than the mud. This early work was followed by a number of attempts to build mathematical models of the displacement process, but these apparently did not match experimental behavior. Some of the assumptions were not strictly accurate, however, and some of the experimental data may have been incorrectly interpreted. One important cause of errors was the assumption that interfacial velocity profiles could be modeled by a single-fluid velocity profile. Early models also assumed that no reaction occurs between mud and cement, which generally is not true. With the increasing use of compatible spacers, however, this assumption may now be valid. The problems of deviated-well cementing were not addressed in earlier modeling problems of deviated-well cementing were not addressed in earlier modeling work but clearly have great significance today. After the poor success of the mathematical approach, a more empirical approach was taken to develop a series of field guidelines. The resultant guidelines included 10-minute spacer-contact time, pipe movement, use of spacers, and further evidence of the benefits of turbulent flow. During this period, the first attempts were also made to investigate the problems of cementing highly deviated wells. The empirical approach proved very valuable in increasing the success rate of primary cementing. The usefulness of these guidelines, however, is limited because they are rather rigid and not easily extrapolated to conditions outside the range of the study on which they are based. They do not allow extreme cases, such as horizontal wells, to be cemented reliably. A more detailed understanding of the mechanisms is necessary if reliable predictions are to be made over a wide range of conditions. predictions are to be made over a wide range of conditions. We therefore studied displacement once again to develop a model capable of predicting behavior under all conditions. Theory Cementing Requirements. To place cement around the entire annulus successfully, three conditions must be satisfied.1. Mud Displacement. The mud gel must be broken down so that mud moves on the narrow side of the annulus, This should be done during mud conditioning before cementing. 2. Yield Stress. The yield stress of each fluid (mud, spacer, and cement) must be overcome to allow fluid to flow into or out of the narrow side of the annulus. 3. Channeling. The velocity of the interface between two fluid in the annulus should be the same on the wide and narrow sides. If the interfacial velocity on the wide side is substantially greater than that on the narrow side, severe channeling will result. The following sections address the problems of displacing one fluid by another in an eccentric annulus. These discussions apply equally to the displacement of mud by spacer and spacer by cement. Remember that almost all annuli are eccentric to some extent. even in vertical wells. Therefore, this analysis can be applied to all well types. Mud Displacement. Lockyear and Hibbert discussed the problem of gelled mud in detail. During mud circulation before cementing, problem of gelled mud in detail. During mud circulation before cementing, the mud gel must be broken down so that the mud is flowing in the entire annulus. With no pipe movement, the only force acting on the gelled mud is the frictional pressure drop. The condition necessary to break down mud gel strength, is ................... .................... (1) where = wall shear stress generated by frictional pressure drop. A simple expression for wall shear stress valid for annuli with diameter ratios approaching unity is = ....................................... (2) where b= width of the annular gap on the narrow side of the annulus and = frictional pressure drop. The frictional pressure drop ideally should be estimated with a hydraulics program that accounts for the effect of standoff because computer simulations have shown that pressure drop is very sensitive to standoff. SPEDE P. 201
Summary This paper describes a 2D hydraulic fracturing simulator that relies on Blot's theory of poroelasticity. The simulator is based on a poroelastic extension of the displacement-discontinuity (DD) method: it uses time and space distributions of impulse-point DD's and impulse-point sources. Because both solutions satisfy Blot's equations, the model fully accounts for the coupling between fluid diffusion and rock deformation. Thus, this model yields a rigorous determination of the influences of fluid leakoff on fracture width and of opening distribution on reservoir pore pressures. Some preliminary results relevant to hydraulic fracturing are then presented. Introduction Models for predicting fluid leakoff, width, and length of hydraulic fractures can be divided into three classes, depending on the complexity of the interaction between diffusion of reservoir/fracturing fluids and deformation of the rock. Class 1-Uncoupled Models. In most hydraulic fracturing models, the stress/ displacement analysis of the reservoir rock is based on the assumption that the rock is elastic. The fracture aperture can be computed from the elastic constants of the rock, in-situ stresses, and pressure distribution inside the fracture. Calculation of the fluid loss to the formation is generally based on Carter's ID diffusion solution, which predicts an instantaneous leakage inversely proportional to the square root of the wetting time. There is no direct interaction between the diffusion and deformation processes, except for a leakoff term in the mass-conservation equations of the fluid-flow analysis inside the fracture. Class 2-Partially Coupled Models. In these models, the stress/displacement analysis is still based on the assumptions of elasticity. The fluid loss is calculated exactly, within the framework of the linear diffusion law, by distributing fluid sources along the fracture. The effect of pore-pressure gradient (caused by leakoff) on rock deformation and therefore on fracture width is accounted for with the concept of back stress. Class 3-Fully Coupled Models. These models include the full range of coupled diffusion/deformation effects predicted by Biot's theory of poroelasticity: sensitivity of the volumetric response of the rock to the rate of loading, pore-pressure change induced by the variation of mean stress, and back-stress effects already accounted for in the Class 2 models. This paper presents the first phase of the implementation of an in-plane, 2D, fully coupled, poroelastic model that uses the boundary-element approach. The general problem is reduced to the investigation of a line fracture along which displacements and fluid flows are discontinuous. These discontinuities are modeled with impulse DD's and source discontinuities, respectively. The interest of the frilly coupled approach resides in the inclusion of the effects of the DD's (fracture aperture) on the stresses and fluid flow in the formation and the effect of the source discontinuities (fluid loss) on the fracture pressure and aperture profiles.
- North America (0.28)
- Europe > Norway > Norwegian Sea (0.25)
Summary. A new motor was developed for use in air/gas drilling operations. Numerical simulation indicated that such a motor was feasible, and a prototype was built. The prototype underwent extensive bench testing and prototype was built. The prototype underwent extensive bench testing and two successful field tests. The downhole pneumatic turbine drilling motor is expected to prove useful in directional drilling operations in areas where formation damage is a problem. Introduction As early as 1938, compressors in a closed circulating loop were used to drill oil wells with natural gas as a drilling fluid, I but the technology of air/gas drilling really began to develop in the early 1950's. The excellent results obtained in drilling seismic shot holes with air led El Paso Natural Gas Co. to attempt gas drilling in the San Juan basin. Using locally available natural gas, they embarked on a highly successful drilling and completions program that led to more than 1,500 gas-drilled wells in the area by 1961. In 1957, Angela aided the growth of the air-drilling industry by devising a reliable technique for estimating the air or gas volume required for hole cleaning. Angel's work was based on the assumptions of a constant friction factor and a minimum gas kinetic energy requirement for lifting the cuttings. More recently, the actual slip velocities of cuttings and the frictional component of pressure losses associated with cuttings transport have been measured, leading to another technique for estimating volume requirements. Ref. 8 discusses such details as surface location elevation and temperatures, misting, and the effect of bit jet size. Introducing air/gas drilling into directional drilling requires the development of a reliable downhole pneumatic drilling motor. The first U.S. patent for a downhole mud motor was filed in 18739; today mud motors are commonly used in the U.S. and internationally. In the Soviet Union, around 80 % of all wells are drilled with downhole motors. A similar development and application of airor gas-driven downhole motors has not yet occurred. A positive-displacement air motor that was field tested in 1960 performed positive-displacement air motor that was field tested in 1960 performed quite well, but it was not reliable because some motor components suffered abrasion. Likewise, air hammers have been in limited use since the early 1960's. Although one test produced rates of penetration (ROP's) at least twice as high as those produced by rotary air penetration (ROP's) at least twice as high as those produced by rotary air drilling, air hammers, used extensively in the mining industry, are not yet common oilwell drilling tools. This paper discusses the design, construction, and testing of a new pneumatic turbine downhole drilling motor. When fully developed, this motor should provide many advantages in both directional and straight-hole drilling. In particular, it will provide the same active trajectory control that is available in mud-drilled wells. Furthermore, use of the new motor in straight-hole drilling will either eliminate or reduce drillstring rotation, which in turn will improve borehole stability, prolong drillstring and bit life, and increase footage. Motor Components and Operating Principles Fig. 1 is a schematic of the downhole pneumatic turbine motor, which is composed of three major parts: the turbine motor module, the gearbox assembly, and the thrust assembly. The turbine motor module and the gearbox assembly are placed in the main housing, which is connected to the thrust assembly. The airway is indicated by the arrows. Fig. 2 shows the turbine motor module, an impulse-type turbine consisting of two major parts: a stationary nozzle plate and a rotor (see Fig. 3). The internal energy of the air stream delivered by the surface compressors is changed to kinetic energy at the nozzle plate. As the high-velocity air flows through the rotor, the kinetic energy of the air stream is changed to kinetic energy of the rotor. This energy is subsequently transferred, by means of the gearbox assembly, to, the main shaft of the thrust assembly and the rock bit. The primary advantage of an impulse-type turbine is a high output capacity (torque and horsepower) obtained in a very limited volume. The gearbox assembly is a sealed unit with its own internal oil lubrication system. It comprises four stages of planetary gears, with a total gear ratio of 168: 1. Energy conservation requires that the reduction in rotary speed be associated with a proportional increase in output torque. Fig. 4 shows the major components of the disassembled gearbox. The thrust assembly transmits the torque from the gearbox output shaft and the weight from the drill collars to the rock bit. It has a large radial bearing near the bit to keep the lower end of the main shaft rotating about its centerline. Above this radial bearing is the main thrust bearing, which carries the entire weight on bit (WOB) when the motor is operating. Above the main thrust bear ing is a series of radial bearings to ensure the concentricity and stability of the main shaft. The top radial-thrust-type bearing allows the main shaft to be rotated while the drillstring is being pulled out of the hole. Also, a bypass valve assembly with an air filter can be placed above the turbine to allow the air flow to be vented from the interior of the drillstring to the annulus of the hole. By venting some of the air to the annulus, the turbine maximum rotary speed is reduced, preventing gearbox damage during operation with little or no WOB. The fully assembled downhole motor is approximately 16 ft [4.9 m] long and 9 in. [228.6 mm] in diameter. The downhole motor described here can drill a hole with a diameter less than or equal to 10 5/8 in. [less than or equal 269.9 mm]. Computer Simulation Numerous factors affect the process of drilling with a downhole motor. Some of these factors are controllable and measurable; others are not. For example, the air flow rate and its inlet pressure and temperature are readily measurable, while the downhole pressures and temperatures are very difficult to measure. These factors, however, can be calculated with a reasonable degree of accuracy and reliability. Nguyen and Johnson both performed theoretical studies in which finite-difference techniques were used to solve the system of equations describing fluid flow and heat transfer in the drillpipe and the borehole annulus. Nguyen used the system approach (constant-mass elements), while Johnson chose the control-volume approach. To arrive at a reasonably simple mathematical description, a number of simplifying assumptions were made. Johnson's major assumptions were as follows:the borehole is vertical, circular, and lined throughout the casing; the drillpipe and drill collars have constant cross-sectional areas and are centered in the hole; transient phenomena related to mass and momentum transfer are negligible; air behaves as a perfect gas with a fixed heat capacity and a viscosity that depends only on temperature; the drilling process results in a constant influx of uniform-sized cuttings (subrounded particles) into the borehole annulus; the motor is an ideal impulse turbine with a given peak horsepower and an efficiency calculated to produce this power; and heat transfer occurs to and from the formation, cuttings, drillstring, casing, and the flowing fluid. SPEDE P. 239
- North America > United States > Texas > Permian Basin > Yeso Formation (0.99)
- North America > United States > Texas > Permian Basin > Yates Formation (0.99)
- North America > United States > Texas > Permian Basin > Wolfcamp Formation (0.99)
- (24 more...)
- Well Drilling > Drillstring Design > Drill pipe selection (1.00)
- Well Drilling > Drilling Operations > Directional drilling (1.00)
- Well Drilling > Drilling Equipment (1.00)
- Well Drilling > Drill Bits > Bit design (1.00)
Summary. This paper presents generalized procedures to interpret pressureinjection and falloff data following cold-water injection into a hot-oilreservoir. The relative permeability characteristics of the porous mediumare accounted for, as is the temperature dependence of the fluidmobilities. It is shown that the saturation and temperature gradients havesignificant effects on the pressure data for both the injection and falloffperiods. The matching of field data to type curves generated from periods. The matching of field data to type curves generated from analytical solutions provides estimates of the temperature-dependentmobilities of the flooded and uninvaded regions. The solutions also may beused to provide estimates of the size of the invaded region, the distanceto the temperature discontinuity, heat capacities, and wellbore-storage andskin effects. Introduction Numerous full-field waterflooding projects are currently under waythroughout the world to improve recovery. In many large oilfields, water injection is initiated during the early stages of reservoirdevelopment. Exploratory wells are tested for injectivity, and injectorsare tested during field operation. If properly interpreted, these testscan give information about the progress of the flood (i.e., frontaladvance), residual oil saturation, the flow characteristics of the virginformation, and near-wellbore damage. In a water-injection well test, the injected fluid usually has atemperature different from the initial reservoir temperature. Duringinjection, both a saturation and a temperature front propagate intothe reservoir. Furthermore, because of differences in oil and waterproperties, a saturation gradient is established in the reservoir. The properties, a saturation gradient is established in the reservoir. The water saturation is highest close to the well and continuouslydecreases with distance from the well. Ahead of this invaded regionis the unflooded oil bank at initial water saturation. For the interpretation of well-test data, the most importanttemperature-dependent fluid property is the viscosity. The viscosity ofboth oil and water may change by an order of magnitude between 50and 572 degrees F, with the major change occuring between 50 and212 degrees F. This temperature effect strongly influences the fluidmobilities, and hence the saturation gradient and the transient pressureresponse. The total fluid mobility changes continuously in theinvaded region and has to be accounted for in reservoir modelingand data interpretation. Many different models have been introduced for the analysis ofwater-injection and falloff tests. Typically, these models neglectthe temperature effects, the saturation gradient, or both. Refs. 2and 3 provide reasonably complete reviews of previous works. For this paper, the most important reference is Fayers'extension of the fractional flow theory of Buckley and Leverett toaccount for a radial temperature gradient in the reservoir. Fayers'work was put into a mathematical framework by Karakas et al.and Hovdan. Hovdan also used this incompressible-fluidssolution to derive a pressure-transient solution for the late stages of acold-water-injection test. Recently, Abbaszadeh and Kamal presented procedures toanalyze falloff data from water-injection wells. Their procedures arebased on analytical solutions not presented in their original paperand include the effect of the saturation gradient in the invaded region. Nonisothermal effects were not considered. In summary, a number of studies pertaining to well-test analysisof injection and falloff tests have been presented. However, none ofthese account for both of the two most important effects in a typicalwaterflood: the saturation gradient and the temperature effect. The principal objectives of this paper are (1-) to derive analyticalsolutions that include the most important effects in a nonisothermalwater-injection/falloff test, (2) to examine the parameters thatinfluence the well injectivity, and (3) to present procedures to obtaindetailed and accurate information about the important reservoir andfluid properties in a waterflood. Specifically, we consider thepressure behavior at the well resulting from the simultaneous flow of pressure behavior at the well resulting from the simultaneous flow of oil and water in a reservoir with a radial temperature gradient. Analytical solutions that account for the effects of temperature andsaturation gradients are derived and discussed. Consequences ofneglecting the temperature and saturation effects are illustrated. Solutions for linear systems, including the effects of linearboundaries in cylindrical reservoirs, were presented by Bratvold andLarsen. Mathematical Model Fig. 1 presents a schematic of the reservoir configurationconsidered. The reservoir is assumed to be cylindrical with the well atthe center. The well penetrates the entire formation thickness, andfluid is injected at a constant rate. The reservoir is assumed to bea uniform, homogeneous porous medium, completely saturated withoil and water. Liquid compressibilities are assumed to be constant, while the viscosities are assumed to be functions of temperatureonly. Neglecting effects of gravity, as well as heat transfer to thesurrounding formation, permits the use of a ID radial model. Injection Period. The transient, nonisothermal two-phase flow of oiland water requires that saturations, pressures, and temperatures bedetermined simultaneously at any time. Furthermore, becausecold-water injection into a hot-oil reservoir is a moving-boundaryproblem, it cannot be solved with standard linear techniques, such as problem, it cannot be solved with standard linear techniques, such as eigenfunction expansion, integral transforms, or Green's functionmethods. To circumvent the problem of simultaneously solving the coupledsecond-order conservation equations, we derive an alternativeapproximate solution to the injection problem using a two-stepprocedure. procedure. Step 1. Assume incompressible fluids. Then use fractional flowtheory to solve the resulting first-order coupled energy- andmass-conservation equations. This essentially amounts to decouplingthe equations for saturation and temperature from the pressureequation. The saturation profile obtained is a Buckley-Leverettprofile including (convective) temperature effects. profile including (convective) temperature effects. Step 2. With the saturation and temperature profiles and themobilities and diffusivities known from Step 1, solve the diffusionequation for pressure by assuming that the fluid compressibilities aresmall and constant. Hence, the pressure distribution in the systemis obtained by superimposing pressure-transient effects on asaturation profile known a priori. Fig. 2 shows an example of a saturation and temperaturedistributions as functions of the similarity transform, ascalculated from the Buckley-Leverett model and includingtemperature effects. Note that the profile exhibits two saturationdiscontinuities. In addition to the discontinuity depicted by the standardBuckley-Leverett theory, the saturation distribution shows asecond discontinuity caused by the step-change in temperature. Themagnitude of the saturation change at the temperature discontinuityis related to the ratio between the mobility ratios in the hot andcold zones. The saturation distribution obtained from a numericalsimulator is superimposed on the analytical saturation profile. Thesimulations were performed with a two-phase, 2D, black-oilsimulator developed by Nyhus that is described later in the paper. SPEFE P. 293
- Water & Waste Management > Water Management > Lifecycle > Disposal/Injection (1.00)
- Energy > Oil & Gas > Upstream (1.00)
Electromagnetic Propagation Logging While Drilling: Theory and Experiment
Clark, Brian (Schlumberger-LWD) | Allen, David F. (Schlumberger-LWD) | Best, David L. (Schlumberger-LWD) | Bonner, Stephen D. (Schlumberger-LWD) | Jundt, Jacques (Schlumberger-LWD) | Luling, Martin G. (Schlumberger-LWD) | Ross, Mike O. (Schlumberger-LWD)
Summary The compensated dual resistivity (CDR) tool is an electromagnetic propagation tool for measurement while drilling. The CDR tool provides two propagation tool for measurement while drilling. The CDR tool provides two resistivity measurements with several novel features that are verified with theoretical modeling, test-tank experiments, and log examples. Introduction The CDR tool is a 2x10(6) -cycles/sec electromagnetic propagation tool built into a drill collar. This drill collar is fully propagation tool built into a drill collar. This drill collar is fully self-contained and has rugged sensors and electronics. The CDR tool is borehole-compensated. requiring two transmitters and two receivers. The transmitters alternately broadcast electromagnetic waves, and the phase shifts and attenuations are measured between the receivers and averaged. Phase shift is transformed into a shallow measurement, Rps, and attenuation is transformed into a deep measurement, Rad. The CDR tool has several new and important features. 1.Rad and Rps provide two depths of investigation and are used to detect invasion while drilling. For example, in a 1-Ohm * m formation, the investigation diameters (50% response) are 30 in. for Rps and 50 in. for Rad. 2. Rad and Rps, detect beds as thin as 6 in., however, these measurements are affected differently by shoulder-bed resistivities and both require corrections in thin resistive beds. Rps has a better vertical response than Rad. Rad and Rps cross over at the horizontal bed boundaries, this crossover can be used to measure bed thickness. 3. Both Rad and Rps are insensitive to hole size and mud resistivity in smooth boreholes. Borehole corrections are very small even for contrasts of 100:1 between formation and mud resistivities. Rugose holes and salty muds together, however, can cause larger errors than indicated by the borehole-correction charts. In these conditions, borehole compensation is essential for an accurate measurement. An extensive research program was conducted to verify these features and to ensure that the CDR tool provides a high-quality log. To achieve wireline quality, the CDR tool's physics was studied thoroughly, and its environmental effects were modeled and experimentally measured. Two theoretical models are used for the CDR tool. The first model treats the tool geometry in detail but assumes a homogeneous medium outside the tool. This model is verified by test-tank experiments and by air measurements. The second model assumes a simplified tool geometry but treats bore-holes, caves, beds, and invasion in detail. This model is used to study environmental effects and to prepare correction charts. Experiments with artificial boreholes, caves, step-profile invasion, and horizontal bed boundaries verify the predictions of the second model. Finally, CDR logs are compared to wireline logs to demonstrate the new features. Theoretical Models The two CDR models emphasize different aspects of the sonde and its environment. The first model accounts for the effects of the antenna recesses and is called the profile-collar (PC) model. It assumes azimuthal symmetry and a homogeneous medium outside the collar (Fig. 1), The azimuthal symmetry reduces the wave equation to a ID integral equation that is solved numerically. This model also accounts for the effect of the finite collar conductivity. The drill collar's varying diameter and finite conductivity produce small but measurable effects in the phase shift (0.1 degrees) and the attenuation (0.2 dB) measured between the receivers. The PC model is used to calculate the transforms from measured phase shift and attenuation to Rps and Rad. Phase shift and attenuation are calculated for many resistivity values and for the dielectric constant. Phase shift and attenuation are monotonic functions of resistivity and are inverted to obtain the transforms. The second model is the uniform-collar (UC) model. It includes the borehole and surrounding rock formations but assumes a constant-diameter drill collar (Fig. 2). The coils are contained in an insulating layer surrounding the collar. The UC model is less accurate than the PC model. The UC model handles up to three beds with arbitrary invasion, a borehole, and washouts. It is used for environmental studies and to generate correction charts for bore-holes, bed thickness, and invasion.
Summary An aquifer-influence function (AIF) can be calculated from a gas reservoir's production and pressure histories. The AIF is unique for an aquifer and can be analyzed to determine aquifer size and other information. Two AIF type curves were developed for aquifers with partially sealing faults and then applied to 32 U.S. gulf coast gas reservoirs. Introduction Aquifers are often associated with gas reservoirs. As gas is produced, water encroaches into the producing reservoir and tends to support the reservoir pressure. Estimating the characteristics of the associated aquifer is vital to proper management of a waterdrive reservoir. Sometimes, the reservoir and associated aquifer are located in a highly faulted geologic setting. Limited geologic and seismic knowledge exists about the location of these faults and about whether they are sealing or partially sealing. This limited knowledge may be supplemented with an analysis of the available reservoir performance data. The rate of water encroachment is determined by the size, geometry, and flow properties of the aquifer and by the pressure drop resulting from gas production. The pressure response of the aquifer may be represented by an AIF-i.e., a unique function of the aquifer that relates the pressure drop per unit rate at the original gas/water contact (GWC) to time for a constant aquifer-influx rate. Many authors have presented models for estimating water influx that can be applied to several different flow geometries (linear, radial, bottom, and edge) and flow regimes, including steady state, pseudo-steady state, and unsteady state. Methods for predicting the performance of aquifer-driven gas reservoirs have also been presented. Hutchinson and Sikora and Katz et al presented methods for calculating AIF's directly from field data. These methods were improved by Coats et al. who used a material-balance program and linear programming to determine AIF's for gas-storage reservoirs. Refs. 19 through 23 have all adapted/modified Coats et al.'s method to compute the optimum AIF for gas-producing reservoirs. Gajdica used an iterative procedure to determine the original gas in place (OGIP) in a dry gas reservoir with an aquifer. Wattenbarger et al. focused on obtaining geologic information about an aquifer from analyzing its AIF. The Cartesian plot of the AIF may be analyzed to determine the aquifer size and the distance to the farthest boundary. Nabor and Barham developed type curves for linear aquifers with limited-, infinite-, and constant-pressure-boundary conditions. Mueller constructed a series of solutions for nonhomogeneous linear and radial aquifers. Stewart et al. considered the effects of a partially sealing fault in an infinite, homogeneous reservoir numerically, and Yaxley considered them analytically for interference testing. Effects of a partially sealing fault in a composite reservoir were also considered analytically by Ambastha et al for transient-pressure testing. These results, however, do not apply directly to aquifers. The current work presents two new type curves for aquifer performance. The type curves are for partially sealing faults in infinite aquifers with linear and radial geometry. The 32 field cases of Gajdica et al. 20 were analyzed for (1) aquifer size and distance to the farthest boundary from coordinate AIF plots and (2) comparison to the two new log-log type curves to determine whether partially sealing faults could be identified.
Summary In this paper, I present the general characteristics of the Gladfelter deconvolution method, which has been used for determining the constant-rate behavior of a system from measured flow rate and pressure. The validity of the method is established for different rate variations and flow geometries (radial, linear, and spherical). The method works well for linear and exponential flow variations and fails for the general flow case. The commonly assumed first semilog straight line resulting from the Gladfelter deconvolution is instead a tangential line parallel to the final semilog straight line owing to a constant-flow-rate period. New solutions are presented for cylindrical, linear (fractured), and spherical wellbore flow geometries with exponential-flow and constant-storage cases. In addition, useful asymptotic solutions are given for the determination of reservoir parameters from the Gladfelter method. Introduction In 1955, Gladfelter et al. stated that the reciprocal productivity index, is a linear function of the logarithm of time for buildup tests, where is the change in shut-in pressure and is the change in afterflow rate. This method, Gladfelter deconvolution, has been used in well testing since 1955 to interpret measured downhole pressure and flow rate. Over the years, a number of studies have appeared on this subject. In terms of dimensionless variables, the Gladfelter deconvolution is the determination of dimensionless pressure at the sand-face, including skin () for the constant-rate case from measured downhole pressure and flow rate. is also called deconvolved pressure, influence function, or unit response behavior. Once is obtained, conventional interpretation methods can be used to determine a well/reservoir system and its parameters. The purpose of this work is to study the general behavior of Glad-felter deconvolution and to explore its application to radial, line, and spherical flow systems.
- Reservoir Description and Dynamics > Formation Evaluation & Management > Pressure transient analysis (1.00)
- Reservoir Description and Dynamics > Formation Evaluation & Management > Drillstem/well testing (1.00)
- Production and Well Operations > Well & Reservoir Surveillance and Monitoring > Production logging (1.00)