Summary We propose a new method to predict the injectivity decline for water-injection wells. Both internal and external filtration are considered in this method. A transition time from internal to external filtration is calculated by use of the trapping efficiency obtained from Stokesian dynamics simulations. The radius and the permeability of the damage zone are calculated by use of a deep-bed filtration model. Equations are derived that allow us to combine the flow resistance in the external and internal filter cakes to obtain the overall decline in injectivity. Introduction A substantial proportion of oil produced in mature fields is produced by waterflooding. Suspended solids in the injected water will cause the injectivity of the injection well to decline with time. Filtration usually can reduce the concentration of suspended solids, but the cost of water treatment must be balanced against the cost of other alternatives, such as periodic stimulation. The prediction of injectivity decline provides a method to specify the water-quality requirements and determine the degree of well impairment. This method allows an operator to design surface facilities to minimize water-injection and well-treatment costs. To predict the loss in injectivity, we need to know the nature of the impairment and its impact on well injectivity. Previous work shows that external and internal filter-cake formation are the most common causes for well impairment (Fig. 1). Various parameters, such as particle/pore size and distribution, injected solids concentration, formation permeability, and fluid velocity, determine whether or not internal or external filter cakes are formed. Several models are available for predicting injectivity decline by use of these parameters and/or core flow experiments. Our objective is to develop a model that properly accounts for both the infiltration of particles and the buildup of an external filter cake at the rock surface. Literature Review Many published studies relate to well impairment or formation damage caused by suspended particles in injected fluids. They generally can be divided into two categories - experimental studies investigating impairment mechanisms by use of core flow tests and models for the prediction of well-injectivity decline. Previously Reported Experimental Observations. Barkman and Davidson1 proposed four mechanisms of well impairment: wellbore narrowing (external cake formation), particle invasion (internal cake formation or deep-bed filtration), wellbore fillup, and perforation plugging. Donaldson et al.2 conducted experiments by flowing colloidal silica suspensions through three types of core plugs taken from Berea, Noxie, and Cleveland sandstones. Their results show that particles initially pass through the larger openings in the core and are stopped gradually by a combination of the effects of sedimentation, direct interception, and surface deposition. The plugging of the core appears to be a combination of internal blockage of pores and cake buildup. They found that the larger particles initiate cake formation. In most cases, the core itself was only partially plugged, but the external cake formed at the face of the core restricted the flow of the suspension. Davidson3 conducted experiments on the flow of particulate suspensions through porous media. A major finding of this work was that the velocity required to prevent particle deposition is inversely related to the particle size. This conclusion is limited to the particular system that the author tested. However, it is the first indication in the literature that there is a relationship between particle movement through porous media and linear-flow velocity. Todd et al.4 conducted particle-plugging experiments on three different core materials. The results of these studies indicate that significant permeability impairment can be caused by inorganic solids, even in very dilute systems. We did not observe the injectivity decline predicted by core studies in the field. The possible reason for this may be the existence of fractures at the wellbore face. Todd et al.5 used aluminum oxide particles in the size ranges 0 to 3, 4 to 6, and 8 to 10 microns. They found the following:The overall damage is related to the mean pore-throat size;The cores damaged with 0 to 3 micron suspensions exhibit damage throughout their entire length; andAs particle size increases, the damage is gradually shifted toward the injection end of the core and to an external filter cake. Vetter et al.6 provided a critical review of past laboratory studies and conducted particle-filtration tests. Their results show that particles of all sizes (from 0.05 through 7 microns) cause formation damage. The larger particles cause a rapid decline in permeability with the damaged region being shallow. Smaller particles (in the submicron range) enter the core and cause a gradual permeability decline. Fluid flow rate is an important parameter for submicron particles from the point of view of permeability impairment. The higher the linear velocity, the greater the depth of particle penetration. Todd et al.7 reported experiments where permeability measurements were made along the length of a pressure-tapped core, 80 mm in length. The core used was a Clasach sandstone with a permeability of 200 to 1000 md and an average porosity of 14.5%. The flow rates were 0.45 to 1.8 cc/sec, particle sizes were less than 3 microns, and particle concentrations ranged from 1 to 15 ppm. The injection time was chosen to be very large (up to 10,000 core pore volumes) to study the importance of long-time injection tests. The results indicate that in-depth invasion occurs for broken-face plugs as compared to external filter cakes, which often occur with sawn-face plugs. Permeability shows a simple semilog decline with gross flow velocity and particle concentration. Smaller velocities and larger particle concentrations result in larger permeability declines. Pautz8 concluded that the 1/3 to 1/7 rule-of-thumb is upheld by coreflooding results when the formation permeability (square root of k) is used to estimate a minimum particle size that would contribute to permeability reduction. When suspended particles in a carrier fluid are flowed through a porous medium, the operative plugging mechanism depends on the characteristics of the particles, the characteristics of the formation, and the nature of the interaction between the particles and the various reservoir materials. The particle/pore size ratio is the most important parameter in the filtration process. It has been found that larger particle/pore size ratios tend to cause rapid, but shallow, damage. Previously reported core experiments do not demonstrate clearly conditions under which an external or internal filter cake is formed. Permeability decline is sensitive to the linear-flow velocity for only very small (submicron size) particles. Particle concentration has less of an impact on permeability impairment for smaller particles. Previously Reported Experimental Observations. Barkman and Davidson1 proposed four mechanisms of well impairment: wellbore narrowing (external cake formation), particle invasion (internal cake formation or deep-bed filtration), wellbore fillup, and perforation plugging. Donaldson et al.2 conducted experiments by flowing colloidal silica suspensions through three types of core plugs taken from Berea, Noxie, and Cleveland sandstones. Their results show that particles initially pass through the larger openings in the core and are stopped gradually by a combination of the effects of sedimentation, direct interception, and surface deposition. The plugging of the core appears to be a combination of internal blockage of pores and cake buildup. They found that the larger particles initiate cake formation. In most cases, the core itself was only partially plugged, but the external cake formed at the face of the core restricted the flow of the suspension. Davidson3 conducted experiments on the flow of particulate suspensions through porous media. A major finding of this work was that the velocity required to prevent particle deposition is inversely related to the particle size. This conclusion is limited to the particular system that the author tested. However, it is the first indication in the literature that there is a relationship between particle movement through porous media and linear-flow velocity. Todd et al.4 conducted particle-plugging experiments on three different core materials. The results of these studies indicate that significant permeability impairment can be caused by inorganic solids, even in very dilute systems. We did not observe the injectivity decline predicted by core studies in the field. The possible reason for this may be the existence of fractures at the wellbore face. Todd et al.5 used aluminum oxide particles in the size ranges 0 to 3, 4 to 6, and 8 to 10 microns. They found the following:The overall damage is related to the mean pore-throat size;The cores damaged with 0 to 3 micron suspensions exhibit damage throughout their entire length; andAs particle size increases, the damage is gradually shifted toward the injection end of the core and to an external filter cake. Vetter et al.6 provided a critical review of past laboratory studies and conducted particle-filtration tests. Their results show that particles of all sizes (from 0.05 through 7 microns) cause formation damage. The larger particles cause a rapid decline in permeability with the damaged region being shallow. Smaller particles (in the submicron range) enter the core and cause a gradual permeability decline. Fluid flow rate is an important parameter for submicron particles from the point of view of permeability impairment. The higher the linear velocity, the greater the depth of particle penetration. Todd et al.7 reported experiments where permeability measurements were made along the length of a pressure-tapped core, 80 mm in length. The core used was a Clasach sandstone with a permeability of 200 to 1000 md and an average porosity of 14.5%. The flow rates were 0.45 to 1.8 cc/sec, particle sizes were less than 3 microns, and particle concentrations ranged from 1 to 15 ppm. The injection time was chosen to be very large (up to 10,000 core pore volumes) to study the importance of long-time injection tests. The results indicate that in-depth invasion occurs for broken-face plugs as compared to external filter cakes, which often occur with sawn-face plugs. Permeability shows a simple semilog decline with gross flow velocity and particle concentration. Smaller velocities and larger particle concentrations result in larger permeability declines. Pautz8 concluded that the 1/3 to 1/7 rule-of-thumb is upheld by coreflooding results when the formation permeability (square root of k) is used to estimate a minimum particle size that would contribute to permeability reduction. When suspended particles in a carrier fluid are flowed through a porous medium, the operative plugging mechanism depends on the characteristics of the particles, the characteristics of the formation, and the nature of the interaction between the particles and the various reservoir materials. The particle/pore size ratio is the most important parameter in the filtration process. It has been found that larger particle/pore size ratios tend to cause rapid, but shallow, damage. Previously reported core experiments do not demonstrate clearly conditions under which an external or internal filter cake is formed. Permeability decline is sensitive to the linear-flow velocity for only very small (submicron size) particles. Particle concentration has less of an impact on permeability impairment for smaller particles.
Summary A decline in injectivity in water injection wells can have a large impact on the economic feasibility of offshore water disposal operations. A case study is presented for an offshore Gulf of Mexico water injection project. Data are presented for five typical offshore wells for which a rapid decline in injectivity was observed due to water injection. The wells were successfully acidized every few months over a period of 2 years. An analysis of the data indicates that in injection wells that are not fractured, such declines in injectivity may be expected even for relatively clean injection water. A comparison of the different completion types indicates that both open hole and perforated completions would have yielded similar results. Cleaner water would have improved the situation but at a substantial cost. Fracturing the injection wells appears to be the only plausible way of substantially increasing the half life of such injectors. In cases where reservoir conditions dictate that the wells not be fractured, the economics of periodic stimulation vs. the cost of installing surface facilities for cleaning up the water should be evaluated using models for injectivity decline. "What if" simulations conducted to study the impact of different process parameters such as injected particle size and concentration, injection rate and reservoir properties were found to be a useful tool in specifying water quality requirements. Introduction The injection of water for pressure maintenance and waterflooding applications or the disposal of produced water is becoming an increasingly important issue in the management of produced fluids. Most mature oil and gas producing regions of the world currently have a high ratio of produced water volume to produced hydrocarbon volume. Large volumes of produced water need to be dealt with for disposal purposes. In offshore operations, the injection of sea water for pressure maintenance or for waterflooding is becoming increasingly common. For these reasons, it is imperative to have reliable models to predict the behavior of water injection wells. Barkman and Davidson1 provided a method for estimating the half life of an injection well when the reservoir properties (kh) and water quality are specified. This method has been extensively applied to evaluate injector performance. Estimates of injector half life based on this model provide us with an order of magnitude estimate of injector half life. More recent models by Van Velzen and Leerlooijer,2 and by Van Oort et al.3 provide us with a basis for using core data in predicting injector half lives. Models for injectivity decline for injection wells with various types of completions such as perforated wells, gravel packed wells and fractured wells have been provided by Pang and Sharma.4,5 Their models allow us to either use core flow test data or empirically derived filtration parameters that can be used in estimating the performance of injectors. In this article, we present a case study of the application of these models to five water injection wells in an offshore environment. The results of the simulation study clearly point to the usefulness of conducting such simulation studies to obtain the optimal water quality requirements which dictate the surface facilities required for maintaining specified injection rates. The case study presents a unique data set for unfractured injection wells in which the performance of the injectors has been carefully monitored over the duration of the project. The quality of the injection water and the procedures taken to treat the water have also been extensively documented. Our main focus in this article is the injectivity decline that is caused by injected fines and their impact on injector performance. Water Injection Project History The goals of this water injection project were to maintain reservoir pressure in the target sand. The target sand is completely unconsolidated and only slightly compacted. Five injectors were drilled and completed below the oil-water contact in the target sand. Each well has a permeability height (kh) product in excess of 53,000 md-ft. (Table 1). Based on reservoir properties and expected injection water quality it was determined that injection rates of 50,000 BWPD could be sustained in radial flow. Avoidance of fracturing was essential to prevent early water breakthrough and to maintain water injection in the target sand. Because the waterflood injection rate target needed to maintain reservoir pressure was essentially the same as the facility injection capacity, out of zone losses could not be tolerated. Well Completions All wells were gravel packed to allow the wells to be produced if necessary to remove near wellbore damage. Wells were underbalance perforated, gravel packed with 20/40 mesh sand and completed with 12 gauge wire wrapped screens. Each well was acidized prior to gravel packing. Significant volumes of viscosified and/or crosslinked materials were employed for fluid loss control after gravel packing operations in all wells. The initial injectivity of all wells was low. Small stimulation treatments were performed with 10% HCl and injectivity increased dramatically but declined rapidly with time. As discussed in more detail later, extensive foam diverted HCl and HCl/HF stimulation treatments were then performed, which resulted in the wells attaining their designed injectivities. However, injectivities again declined more rapidly than expected. Waterflood Facilities Seawater is taken from 150 ft subsea, deoxygenated and filtered through primary multimedia filters and secondary cartridge filters. Five µm filters were used in some wells for a period of time but were found to be expensive to operate since they had to be replaced twice a month. In all cases these filters were replaced with 10 µm filters. As discussed later, the resulting change in water quality was clearly reflected in the well injectivity. Oxygen is taken to 200 ppb by countercurrent gas stripping and chemically scavenged to <10 ppb. A combination of continuous and batch treatments with sodium hypochlorite is employed to control bacteria. A scale inhibitor is added because the water shows a slight calcium carbonate scaling tendency.
Injectivity Decline in Water Injection Wells: An Offshore Gulf of Mexico Case Study Mukul M. Sharma, SPE and Shutong Pang, The University of Texas at Austin, Kjell Erik Wennberg, SPE, IKU Petroleum Research, Lee Morgenthaler, SPE, Shell Oil Co. Abstract Injectivity decline in water injection wells can have a large impact on the economic feasibility of offshore water disposal operations. A case study is presented for an offshore Gulf of Mexico water injection project. Data are presented for five typical offshore wells for which a rapid decline in injectivity was observed due to water injection. The wells were successfully acidized every few months over a period of two years. An analysis of the data indicates that in injection wells that are not fractured, such declines in injectivity may be expected even for relatively clean injection water. A comparison of the different completion types indicates that both open hole and perforated completions would have yielded similar results. Cleaner water would have improved the situation but at a substantial cost. Fracturing the injection wells appears to be the only plausible way of substantially increasing the half life of such injectors. In cases where reservoir conditions dictate that the wells not be fractured, the economics of periodic stimulation versus the cost of installing surface facilities for cleaning up the water should be evaluated using models for injectivity decline. "What if" simulations conducted to study the impact of different process parameters such as injected particle size and concentration, injection rate and reservoir properties were found to be a useful tool in specifying water quality requirements. Introduction The injection of water for pressure maintenance and waterflooding applications or the disposal of produced water are becoming increasing important issues in the management of produced fluids. Most mature oil and gas producing regions of the world currently have a ratio of produced water volume to produced hydrocarbon volume in excess of 25 to 1. This indicates that large volumes of produced water need to be dealt with for disposal purposes. In offshore operations, the injection of sea water for pressure maintenance or for waterflooding is becoming increasingly common. For these reasons, it is imperative to have reliable models to predict the behavior of water injection wells. Barkman and Davidson provided a method for estimating the half life of an injection well when the reservoir properties (kh) and water quality are specified. This method has been extensively applied to evaluate injector performance. Estimates of injector half life based on this model provide us with an order of magnitude estimate of injector half life. More recent models by Van Velzen and Van Oort provide us with a basis for using core data in predicting injector half lives. Models for injectivity decline for injection wells with various types of completions such as perforated wells, gravel packed wells and fractured wells have been provided by Pang and Sharma. These models allow us to either use core flow test data or empirically derived filtration parameters that can be used in estimating the performance of injectors. In this paper, we present a case study of the application of these models to 5 water injection wells in an offshore environment. The results of the simulation study clearly point to the usefulness of conducting such simulation studies to obtain the optimal water quality requirements which dictate the surface facilities required for maintaining specified injection rates. The case study presents a unique data set for unfractured injection wells in which the performance of the injectors has been carefully monitored over the duration of the project. The quality of the injection water and the procedures taken to treat the water have also been extensively documented. Our main focus in this paper is the injectivity decline that is caused by injected fines and their impact on injector performance. Water Injection Project History The goals of this water injection project were to maintain reservoir pressure in the target sand. The target sand is completely unconsolidated and only slightly compacted. Five injectors were drilled and completed below the oil - water contact in the target sand. Each well has a permeability height (kh) product in excess of 53,000 mD ft. (Table 1). P. 341^
The performance, i.e. the half-life of water injectors, is very important to the economics of water injection projects. A simulator is presented to predict the decline in injectivity of water injection wells. Equations are presented that allow us to model open-hole and perforated completions as well as fractured wells. The mechanism of injectivity decline can be determined by comparing core flow test data with one of four type curves. The transition time is used to determine when external filtration becomes dominant so that the appropriate models can be used for early and late time. The simulator can be used whether or not core flow data is available. Empirical equations are provided to estimate the filtration parameters if core flow test data is not available. The simulator is easy to use, has interactive input and output and is available for PCs and Unix workstations. Examples are provided to show how the simulator can be used to predict the performance of injectors in a variety of cases. The simulator predictions are compared with experimental data. The simulation results are then used to suggest the optimum water quality that will minimize the total cost of water treatment and periodic well stimulation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.