Darcy's law can not describe fluid flow accurately when the flow rate is high. In most cases in the recovery process, fluid flow is governed by Darcy's law. But when the flow rate is very high, for an instance, near the wellbore, Darcy's law is inadequate to describe fluid flow.
In 1901, Forchheimer put forward a classical equation, known as the Forchheimer equation, to make up the deficiency encountered by Darcy's law at high flow rates. He added a non-Darcy term into the Darcy flow equation. The non-Darcy term is the multiplication of the non-Darcy coefficient, fluid density, and the second power of velocity. One of the most important aspects in determining the non-Darcy effect is to estimate the non-Darcy coefficient as accurately as possible.
In this paper, theoretical and empirical correlations of the non-Darcy coefficient in one-phase and multi-phase cases in the literature are reviewed. Most researchers have agreed that the non-Darcy effect is not due to turbulence but to inertial effect. The non-Darcy coefficient in wells is usually determined by analysis of multi-rate pressure test results, but such data are not available in many cases. So, people have to use correlations obtained from the literature. This paper summarizes many correlations in the literature, and will provide a good reference for those who are interested in the investigation of the non-Darcy effect in the recovery process.
In most cases (not near the well-bore) in recovery processes, the flow pattern is governed by Darcy's law, which describes a linear relationship between pressure gradient and velocity as follows,
where u is superficial velocity, K is permeability, p is pressure, µ is viscosity, and x is dimension in x direction.
Forchheimer1 found that the pressure gradient required to maintain a certain flow rate through porous media was higher than that predicted by Darcy's law. He added a non-Darcy term to Darcy's law to account for this discrepancy, and the flow equation became
where ? is fluid density, and ß is called the non-Darcy coefficient in this paper. From equation (2), we see that the non-Darcy term is a multiplication of the second power of velocity, fluid density, and ß. There have been many names for ß. ß was called: the turbulence factor by Cornell and Katz,2 and Tek et al.;3 the coefficient of inertial resistance by Geertsma,4 and Al-Rumhy et al.;5 the velocity coefficient by Firoozabadi;6 the non-Darcy flow coefficient by Civan and Evans,7 Liu et al.,8 Grigg and Hwang,9 Narayanaswamy et al.,10 and Li et al.;11 the Forchheimer coefficient by Ruth and Ma;12 Inertial Coefficient by Ma and Ruth;13 the beta factor by Milton-Taylor;14 the non-Darcy coefficient by Thauvin and Mohanty,15 Cooper et al.,16 and Li et al.11 Equation (2) is called the Forchheimer equation by Ruth and Ma,12 Milton-Taylor,14 Ma and Ruth,13 Civan and Evans,17 Thauvin and Mohanty,15 Coles and Hartman,18 Cooper et al.,16 and Li et al.11
When flow rate is very high, Darcy's law is not adequate to describe flow pattern. High-velocity gas flow occurs in the near-well-bore region and condensate reservoirs. Non-Darcy effect is important in these regions according to Kalaydljian et al.19
Numerous waterflooding projects are under way throughout the world for increased recovery. Water injection tests of oil zones are frequently undertaken during the planning phase of waterfloods. Analysis of the bottomhole pressure data recorded during these tests not only provides similar information to that obtained from production tests concerning the well and the reservoir characteristics but also allows the mobility ratio between the injected and resident fluids to be determined.
Conventionally, pressure fall-off test data is analyzed using semilog plot of bottomhole pressure versus time. This paper is the extension of the Tiab's Direct Synthesis Technique10-15 to pressure injection and Fall-off tests in water injection wells.
Direct synthesis is a transient pressure analysis technique10-15, which uses log-log plot of pressure and pressure derivative vs. time. Thus, different straight line portions indicating different flow regions are directly analyzed. Direct synthesis is very useful in conditions of short and early time pressure data missing tests. It also verifies the results since it uses more than one equation for the estimation of reservoir parameters such as permeability, wellbore storage coefficient, and skin factor.
Finally, field examples of pressure falloff analysis are presented to illustrate use the direct synthesis and results are compared with those from type curves and conventional semilog analysis.
Traditionally water flood schemes have been implemented later in the life of the field following primary depletion. Now, such schemes are often considered during the initial development of a field. The economic viability of many fields depends upon successful implementation of water injection at early stage. Injection tests are, therefore, performed on appraisal wells drilled prior to the decision to develop the field. These tests are designed to assess both the efficiency of the filtration equipment and the injection characteristics of the formation. Operational and the cost considerations dictate that the maximum possible information be derived from these tests, which may be few hours of duration.
Analysis of the pressure Falloff and injectivity tests has been discussed at considerable length in the literature. The pressure buildup during injection period, however, has received relatively little attention. The main reason is that falloff tests match to the pressure buildup test in production wells, which is easy to analyze. Furthermore, the injectivity test is mathematically difficult to handle due to moving boundary, the flood front.
Ras Budran filed is a massive Nubian sandstone reservoir compartmentalized by major faults acting as partial barriers. Vertical communication is impeded by hydraulically sealing shale layers in the development of three main pressure regimes. A combined water injection and aquifer support the pressure. Pre-mature water breakthrough has been occurred in the middle of oil leg, which in turn limit the corrective action of the isolation of the watered out zones.
3D- geological and reservoir simulation model was constructed, based upon new reservoir characterization with the objective of improving the vertical definitions within the reservoir. The reservoir properties generated by deterministic interpretation for the new micro-zones.
This paper presents the approach taken to match 17 years of production data in particular, via aquifer definition and fault communication.
The history-matched model was then used to confidently check the production and injection well pattern and performance. The matched model then used to investigate the viability of infill wells to improve drainage pattern and sweep efficiency meanwhile increasing ultimate recovery.
Ras Budran field (R/B) is located at the eastern coast of the Gulf of Suez area (Fig 1). The field was discovered in April 1978 and production started in Feb., 1983. Production is maintained by gas lifting while the pressure is supported by a combined water flooding and limited aquifer drive.
There are 17 producing wells and 4 injectors over 3 offshore platforms. The field is relatively deep reservoir with the original oil water contact at 12350 ft-tvdss. Heavily under saturated reservoir with initial reservoir pressure of 5632 psia and the bubble point pressure 1200 psia. The structure contour map Fig. (2) shows the reservoir complexity.
The reservoir is massive sandstone and the macro layers were defined from top to bottom as follows; Raha, Nubian III, IIB, IIA and Unit I. Unit IIA has a shale/sand streaks that work as a vertical hydraulic barrier between the upper and lower units. Lower Units of block, which called Unit I (LA), Upper units of block A (UA), Upper Units of block B (UB) and Upper Units of block C (UC).
Fluid flow in the reservoir is directed from unit I to the juxtaposed upper reservoir units supported by water injection in unit I from injector A3b and direct aquifer support. Fluid flow in upper units of block A is only supported by the water injection through injectors A2, A1 and A9 respectively. Fig (3) illustrates the main cross section for the field.
Due to the nature of Ras Budran reservoir and its dipping structure, peripheral injection pattern was proposed and implemented in the original field development plans in October 1985 through two wells A2a (Unit IIB+III+R), and A3b (Unit I). In January 1990, the system was upgraded with another injector A1 (Units IIB+III), to replace the poor injector A2a and finally, the last injector A9a (Units IIB+III+R) was brought on line in April 1992.
The difference between formation and injection seawater salinity was used as a tool to monitor the flood front and the interblock communication. The field has already produced 80% from the estimated reserves.
Because it will not be possible to capture the dynamics of a water flood project, in particular with high mobility ratios as in the case of Ras Budran, with single cell calculations. Reservoir simulation modeling is essential in order to optimize injection/production well patterns and to optimize sweep efficiency via the injection distribution.
To date, field wide CO2 flooding is continuous or WAG injection processes. In individual wells a huff'n'puff scheme has been used where CO2 is injected into a well followed by a shut-in period; after a predetermined time, the well is produced. Each of these methods have practical, technical, and economical limitations. This paper describes a new CO2 injection method that combines all three of these methods but excludes water injection.
Continuous CO2 may be optimal in reservoirs that are not conducive to water injection and do not have CO2 mobility control problems such as early breakthrough. However, continuous CO2 requires a large initial CO2 volume that may be unavailable and relatively expensive. WAG is very common with variations of different ratios of water to CO2 volumes and tapered ratios. WAG reduces mobility problems and thereby improves areal sweep efficiency; CO2 purchase expenses may be lower due to the requirement of lower CO2 volumes. Unfortunately, water injectivity following CO2 injection may decrease, and the water increases lift and water handling expenses. Huff'n'puff is effective similar to a near wellbore stimulation, but may not realize the benefits of a full field injection program. Also, this adds the complication of injection and production capabilities required in the same well. This may be impossible for some wells on artificial lift.
The proposed method is to inject CO2 and shut in the well similar to the huff and puff, but instead of producing the well inject CO2 again. This process is a cyclic process like WAG, and shares the variables of the duration and volume of CO2 injection and the duration of the shut-in period. This method eliminates injectivity and production expenses associated with water. Mobility is controlled by near wellbore achievement of miscibility by the increase in mass transfer between the CO2 and oil during the shut-in portion of the cycle; i.e. a smaller portion of the reservoir has relatively low viscosity, injected CO2 due to miscibility between oil and gas occurring nearer the injection well.
The use of CO2 as a method of enhanced oil recovery has been studied since the early 1950's1 and its use grew signficantly in the 1970s and 1980s.2 It is used as an enhanced oil recovery (EOR) process where it is injected following either natural drive or waterflooding in order to recover additional oil.
During the life of an oil reservoir, production is usually carried out by primary recovery, secondary recovery, then enhanced oil recovery (EOR). Primary recovery uses the natural energy present in the reservoir to displace oil to the wellbore and includes solution-gas drive, gas-cap drive, natural water drive, fluid expansion, rock expansion, and gravity drainage. Secondary recovery processes add energy to the reservoir by injecting water or immiscible gas and displacing oil to an adjacent producing wellbore. However, because immiscible gases have more mobility problems than water, the secondary recovery stage is almost always carried out with a waterflood. EOR processes are carried out by gas injection, liquid chemical injection and/or the addition of thermal energy. Gas injections are considered EOR processes if the significant recovery mechanisms are not exclusively those of immiscible frontal displacement with high interfacial tension (IFT) relative permeabilities, but include such mechanisms as oil swelling, oil viscosity reduction or favorable phase behavior. The gases used in EOR processes include hydrocarbon gases, CO2, nitrogen, and flue gas. Liquid chemicals include polymers, surfactants, and hydrocarbon solvents. Thermal methods include steam drive, cyclic steam flooding, hot water flooding, and in situ combustion.3
The use of methanol in fracturing is not new technology. Its origin and application in hydraulic fracturing can be traced back to the 1960's. However, advances in fracturing fluid chemistry and breaker technology have improved dramatically creating new systems such as crosslinked methanol, the subject of this paper. Methanol has been utilized in the well stimulation industry for many years to take advantage of its low surface tension properties and miscibility with various formation fluids. This paper will discuss advances in crosslinked methanol system technology, system compatabilities, field application and practical safety precautions. The fluid system rheological properties and proper breaker application will be discussed as well.
In recent years, the industry has become more aware of the challenges associated with formations recognized as "water sensitive". These issues can be associated with mobile and swelling clays, or undersaturation of the formation, otherwise known as aqueous phase trapping. The paper will discuss these issues as they pertain to proper fluid selection and more specifically, the selection of crosslinked methanol as a fracturing fluid.
Formation properties of zones treated with crosslinked methanol will be presented as well as post fracture treatment results.
Crosslinked anhydrous methanol is an excellent fracture fluid system comprised of approximately 96%-100% pure methanol developed to stimulate low permeability water sensitive formations. In the late 1960's, fracturing fluid systems incorporating carbon dioxide (CO2) and methanol saw wide usage and some success in fracturing applications. These systems were used in an attempt to place proppant into a hydraulically induced fracture without the potential damage associated with water on "tight" formations displaying water sensitive tendencies. In the early 1970's, certain patents were awarded for a methanol fluid system termed "Vapor Frac," which consisted of transporting propping agents in gelled alcohol and CO21. This system was used for several years with varying degrees of success due to limited sand transport capabilities and poor fluid leak-off properties. In 1980, the first crosslinked methanol system was introduced to the industry. However, at this time the system incorporated 80% pure methanol and 20% water. While this system was compatible with CO2, it did not fully address the water sensitivity issue. By the late 1980's the first pure (96% methanol plus) methanol fluid system was developed and introduced to the industry, and saw wide use throughout Western Canada.
In late 1992, additional opportunities to apply this technology presented themselves and research was undertaken to apply new fluid technology to crosslinked methanol. The system was introduced in Argentina; originally, to reduce treatment cost relative to CO2 foam systems prevalent at that time. Since then, more than 200 jobs have been pumped2. The accumulated experience gained in Argentina further enhanced and expanded the original intended applications.
In late 1998, depressed crude oil prices and a subsequent decline in drilling operations in the Permian Basin caused pumping service companies to seek applications for this improved crosslinked methanol fluid system in existing producing wells. The hope was to create a stimulation market in gas wells perceived as water sensitive, depleted or damaged to a point that conventional stimulation was not believed to be an option. It is in this application that crosslinked methanol has seen its greatest success in the Permian Basin.
Rivera de la Ossa, Juan E. (ECOPETROL-ICP) | Correa Castro, Juan C. (Universidad Industrial de Santander) | Mantilla Uribe, Cesar A. (Universidad Industrial de Santander) | Duarte Prada, Cesar Augusto (E.O.S. Ltda)
It nitrogen injection has been proven as a competitive alternative forvolatile oil reservoirs lately found in Piedemonte area of Colombia, takinginto account the reservoir fluid nature and the reservoir conditions. Alaboratory study was carried out by the authors in the Enhanced Oil RecoveryLaboratory of Colombian Petroleum Institute.
Three corefloods were carried out using Berea sandstone and volatile oilsamples from the Piedemonte area and nitrogen as displacing agent. One of thesecorefloods was conducted in a long Coreflooding apparatus, and the other ones,in a short Coreflooding apparatus connected with a slim tube in a order toachieve miscibility conditions. The main variable studied was the flow rate,which represents three different flow regimes. The scaling criteria taken was adimensionless group, which relates viscous and gravity forces, called gravitynumber.
Displacement efficiencies determined were around 50%, 65% and 80% for eachone the tests. The results of these corefloods were compared to determine thebest scheme of nitrogen injection. Results have successfully confirmed theeffectiveness of nitrogen as displacing agent for this kind of oils and thebetter swept efficiency given by the equilibrium between the governingforces.
The volatile oil reservoirs lately found in the Piedemonte area of Colombiaare quite complex not only in geology but in the fluids composition also. Thatis why, several analyses have been done to select the best explotation,production and commercialization strategy.
Reservoirs in Piedemonte area are more than 15000 ft deep, with pressuresand temperatures greater than 5500 psi and 250°F. Thermodynamic and phasesbehavior of this reservoirs are also unique in the world.
Fluids behavior have a great importance in the hydrocarbons volumeestimation and therefore in the field development. There is gas reinjection inthe field to keep the initial reservoir pressure, in an effurt to maintain asingle phase fluid and to improve the recovery. Reinjection gas come from afirst stage separator with high methane content.
When observing different studies, made by investigators of world-widerecognition in the area of volatile oil deposits, that propose the nitrogeninjection like an alternative able to compete with the gas reinjection in thistype of deposits, since the characteristics that favor the reinjection of thegas are also propitious for the nitrogen injection, the investigation groupdecided to orient its efforts in the investigation with nitrogen as a new formof profitable operation.
The replacement of the natural gas by nitrogen as injection gas bringsadvantages that are according to the plans of economic development of thecountry.
A series of displacements were doneto investigate and to quantify the effectof phases behavior and porous media on the efficiency of sweeping withnitrogen. An overhaul about experiences of operation with nitrogen in differentoil fields including the facilities required for a project of nitrogeninjection, was initially done.
Barnhart Field was discovered in 1941 on the Ozona Uplift, southeastern Reagan County, Texas. The field produces from the Ellenburger Group on approximately 10,000 acres that can be divided into a southern part on University Lands and a northern part on private land. Two pilot waterfloods were attempted; neither was successful. The field was essentially abandoned in 1976 after producing more that 16 million BO. The structure of the field is an asymmetrical anticline trending northeast - southwest with the crest on the northern part of the field. Structure alone does not appear to control production because the southern part produced slightly more oil than the northern part. Insoluble residue analyses and lateral-log correlations indicate the Ellenburger in the northern part of the field was more deeply eroded than in the southern part. The Ellenburger Group, in this field, can be divided into three pay zones, the Upper Honeycut, the Lower Honeycut and the Gorman. The Upper Honeycut produces only in the southern part of the field and may have the best developed porosity in coarse crystalline dolomite. An isopach of the Honeycut pay zones shows fair agreement between thickness and production. The Gorman pay zone does not correlate well indicating that production cannot be explained by gross isopachs alone.
The town of Barnhart was established in 1910 along the Kansas City, Mexico and Orient Railroad. The land agent for the railroad at the time was William F. Barnhart and the town was named in his honor4. On September 9, 1941, the Amerada Petroleum Corp., University RA #1 made an Ellenburger discovery. Located in southeast Reagan County, Texas, on the Ozona uplift of the Permian Basin about seven miles east of the town of Big Lake, Amerada initially wanted to name the field Big Lake, but a field already existed by that name. Eight miles to the east was the town of Barnhart and Amerada choose that name.
By early 1953, eighty-three producers and five dry holes (Figure 1) had been drilled to complete the initial development of the approximately 10,000-acre field. Field spacing was 160 acres but was later reduced to 80 acres in many parts of the field, most noticeably on University Lands. Approximately half the productive acreage was on University Lands and half on private lands.
Although many Ellenburger fields have strong water-drive reservoirs, particularly those located on the Central Basin Platform, the Barnhart Field is solution-gas drive. A production curve from the University Lands portion of the field shows field volumes from 1941 through 1975 (Figure 2). The hyperbolic nature of the curve is one indication of solution-gas drive. Another indication is that water production remained low over the life of the field and the wells did not "water out." Finally, GORs increased as reservoir pressures declined. Although several wells were reclassified from oil to gas in the 1970s, the presence of a gas cap was not established.
Two pilot waterfloods were conducted from 1960 through 1971. The first was a one-well dump-flood, in which San Andres water was allowed to gravity inject into the open Ellenburger perforations. Poor records were kept for this flood, but based on water levels in the dump-flood well, it appears to have reached hydrostatic equilibrium early in the flood and did not inject much water into the Ellenburger1. The second waterflood was a five-spot in Block 48, Sections 3 and 4, University Lands. Again, insufficient water was injected to achieve fillup. No response was observed from either flood1.
The field was essentially abandoned in 1976 when most of the wells were plugged for the salvage value of the casing. In the late 1970s, several wells were drilled or recompleted in the field in an attempt to locate bypassed oil. Since none of these wells initially produced much more than 20 BOPD or cumulatively produced much oil, they are considered unsuccessful.
Many oilfield service companies as well as oil producers have manually tracked assets over the years for a variety of reasons. The service companies have tracked assets such as pumps, packers, or other products to assist in R&D efforts. Being able to collect the data and compile statistics on run times and component failures enables the service companies to evolve and improve current products as well as design new products. From a producers perspective, compiling the same or similar data allows for the development of "best practices" in operating procedures and processes. Software products developed to address these needs have evolved with time to provide more functionality. However, many systems implemented to handle the total process of data acquisition, warehousing, querying, and reporting to achieve improved operating results have become more difficult and expensive to support than the value added.
The value of the information has not diminished, but increased due to the fluctuations in oil prices and the continuing efforts to reduce lifting costs through design and process enhancements. The recent development of a web based tracking system incorporates workover management, downhole equipment, and chemical usage while enabling the operator and service provider the ability to easily enter and access the data. The system reduces the problems of database synchronization, multiple entries of the same data, and provides a common means through the Internet to interface with the information. The system links the operator in the field, the service company providing equipment or chemicals, and the district office together through a common database that each has access to.
The system allows a technician in a pump shop, workover foreman in the field, or chemical sales person to easily enter data into the system using a laptop computer or touch screen technology. The data is brought back to the service provider's local office and is accessible to the operator through the Internet. Wells, well equipment, and equipment components can be tracked for run life and root cause of failure. The operation becomes an information network that uses the same data to accomplish different tasks but with a common objective of reduced costs and improved profitability.
Water injectivity decline is a very common phenomenon in waterflooding fields. Most of the previous analyses were focusing on water injectivity decline due to the migration of suspended particles in injection water or the injection water/reservoir fluid incompatibility. However, in some unconsolidated formations, another possible mechanism for water injectivity decline is sand mobilization, which means sand particulates separate from rock matrix and move into deep formation. This kind of injectivity decline is controlled by the operation condition of water injection such as injection pressure and injection rate. In this paper, a mathematical model is proposed to simulate the process of sand mobilization and the resultant water injectivity decline.
The mathematical model is derived based on material balance for water, sand particulate, and rock matrix. Also included in this model are particulate generation and deposition constitutive laws, and permeability-porosity correlation. Finite difference scheme is introduced to discretize the partial differential equations and the finite difference equations are solved implicitly through iteration. Sensitivity analysis is performed to study the effects of various factors on water injectivity decline and strategies for managing efficient water injection are proposed through the analysis.
Numerous researches[1-13] have been done in oil/gas well sand production. Because water injectors are not generally back produced and as a result very few researches were done in the past on injector sanding. Literature survey indicates that only a couple of papers[14-15] were published regarding this topic. Despite that, it does not mean sanding in water injectors is not a problem, instead, it can cause the injectivity decreases dramatically. As stated in reference, the injectivity of a well operated by Statoil in the Norwegian Sea decreased from 8000 m3/d to 0 m3/d in just half an hour which is tied to formation failure caused by the pressure waves generated during the sudden shut down of the pumps. From this single example, we can see that how bad it can be in water injectors once sanding occurs. Because of this, researches on sanding in water injectors are of the same importance as those in producers.
Santarelli et al presented a field case study on a reservoir operated by Statoil in the Norweigian Sea concerning the injectivity decline of water injectors. In this paper, it is believed that sanding is caused by the following reasons: 1) During well shut-in, the rock around the well is too weak to sustain the stresses and fails. 2) Because of the reservoir permeability heterogeneity, the wells are cross-flowing during shut-in and cause sand production in front of the perforated intervals. 3) The produced sand is not able settle down in the rat-hole before injection restarts and hence plugs the perforation tunnel. 4) As a result of the water hammer effect caused by well shut-in, the formation already weakened by sand production undergoes liquefaction that triggers large amounts of sand to be released in the well and hence killed the injectivity. Morita et al provided guidelines for completing water injection wells.
The cessation of CO2 injection may be the result of poorreservoir response, low oil prices, or near the end of a CO2project. The focus of this paper is to describe the effect ofCO2 injection curtailment on oil recovery and production in atechnically successful CO2 flood of an oil reservoir. TheCO2 curtailment is sustained over a significant period of time whilewater is injected. Following the curtailment, CO2 is againinjected.
Early EOR literature covers CO2 and miscible gas injection in thepresence of very high water saturation. The key issues are phasebehavior, surface tension, viscosity, and oil stripping, which are described interms of curtailment and the subsequent affect on oil production.
A survey of pertinent literature has been conducted to infer the impactCO2 curtailment has on reservoir performance. Preliminaryfindings are that the curtailment of CO2 leads to a permanent lossin reserves and production rate. Scenarios that may alleviate theselosses such as reduced CO2 injection instead of complete curtailmentare addressed.
A CO2 flood may become uneconomic due to high operation costs orlow oil prices. Due to the expense of purchasing CO2 forinjection, CO2 purchases may be halted until a more favorableeconomic situation exists. The effects of curtailing CO2injection, replacing it with water injection for a significant period of time,and then starting CO2 at a later time is the subject of thispaper.
An analogy to curtailment is injecting CO2 at initially highwater saturations, which has been studied extensively in the literature due toearly concerns of the effectiveness of CO2 to mobilize waterfloodresidual oil. The key issues are wettability, solubility ofCO2 in water, phase behavior (miscibility), diffusion, dispersion(viscous fingers), and oil bank formation.
In an attempt to understand the technical aspects of CO2curtailment, field case publications were intentionally not included.While this work has not been previously published, it was first presented atthe 1999 CO2 Conference in Midland.