This work shows experimental values of compressibility factors for CO2-hydrocarbon mixtures in different proportions. The measurements were done in a Ruska supercompressibility cell varying the pressure (200 - 2500 psia), the temperature (100 - 200 deg. F) and the molar percentage of CO2 (85 - 100%). With the experimental data, generalized Z (CO2-Hydrocarbons) were plotted as a function of Psr (pseudo reduced pressure) and Tsr (pseudo reduced temperature).
The equation state of Peng-Robinson (EOS-PR), Starling (EOS-S) and Redlich-Kwong-Soave (EOS-RKS) were tested to simulate the experimental data, and it was found that the EOS-S and EOS-PR are the ones that predict the mixtures' compressibility factors studied with greater precision. precision
The carbon dioxide (CO2) has multiple uses in the oil industry. Since the 60s, it has been used for the stimulation of oil wells, in the enhanced recovery of oil (EOR) through the miscible and unmiscible displacement of oil at high pressure and moderate temperature, and recently, research is being carried out for its use in the refining of oil through the super-critical extraction of hydrocarbons. In these processes, the CO2 is found mixed with paraffinic hydrocarbons in proportions. that reach paraffinic hydrocarbons in proportions. that reach up to 10% molar. From here, the importance to know with exactitude the compressibility factors of CO2 either pure or mixed with natural gas at moderate and high pressures. This work presents experimental values for CO2 compressibility factors and its mixtures with hydrocarbons, and theoretical and semi-empirical methods are evaluated to allow determining these compressibility factors with a certain degree of exactitude when no experimental information is available.
This paper is concerned with the role of engineering input to, and interaction with, algorithms used in reservoir history matching. The underlying iterative cycle common to all techniques allows considerable scope for such input; however, for maximum benefit to result, it is important that the information accurately reflects engineering experience, and can be suitably quantified for use in algorithms.
The discussion is illustrated with some salient points of examples based on one technique points of examples based on one technique which uses a stochastic framework to accommodate explicitly the uncertainty in such essentially deterministic problems.
A challenging area of study in reservoir engineering is the class of problems under the general heading of history matching; this is illustrated by the diversity of possible solution techniques that are considered in recent and current research. For instance, a recent discussion of a history matching competition outlines a range of techniques, from purely deterministic to wholly subjective.
This paper is based on the central thesis that, given the range and potentially complicated nature of history matching problems, it is sensible for solution techniques to incorporate information from the reservoir engineer when possible, and to allow an experience user possible, and to allow an experience user considerable control over various aspects of the matching process.
We then argue that the information generally available from the user is geostatistical in nature, and discuss a framework to make use of this information. We then consider the role of continuing engineering interaction with solution algorithms, and demonstrate that the framework outlined has sufficiently flexibility to reflect changes in engineering interpretation of the problem under study. We illustrate this part of problem under study. We illustrate this part of the discussion with salient points from examples.
The paper concludes with a brief summary of current work and extensions to the methodology described here, and gives an outline of areas for further work on this topic.
Since May 1991 a POSEIDON pump of the rotodynamic type has been running on the SIDI EL ITAYEM field in TUNISIA, logging more than 3500 operating hours (end of November 91) without a major problem. This pump is one of the practical results of the POSEIDON project, launched in 1984 by TOTAL, IFP and STATOIL, to develop the concept of multiphase pumping: the final target being a subsea multiphase boosting station.
Prior to the Tunisian test, this pump had been Prior to the Tunisian test, this pump had been tested satisfactorily on the two phase loop of IFP at SOLAIZE (FRANCE), and it was this which decided the partners to send the prototype, without any modification, to the real active field test rig specially prepared to accommodate multiphase equipment.
The choice of the multiphase pump type is the result of comprehensive knowledge and comparative studies of the different available or described pumping systems at commencement of the project pumping systems at commencement of the project The partners, not being bound to any particular pump manufacturer, only wished to select or to pump manufacturer, only wished to select or to improve a known product in order to have a tool available for their future subsea developments. In the following paper we will describe the different types of possible multiphase pumps and the reasons that lead the partners to the rotodynamic principle type. We will then present the POSEIDON pump, the tests performed on a diphasic loop to ascertain its behaviour with any composition of fluid, then the tests conducted in real field conditions at SIT and an intermediate conclusion about these operations that mark the end of the R and D phase.
2. MULTIPHASE PUMPS
With the current low price for produced oil and the increasing cost of offshore operations it makes sense to use the declining fields separation capacities to receive the effluent coming from the smaller or marginal fields which do not deserve a conventional development.
Also onshore fields located in remote areas or locations where it becomes more difficult to flare the produced gas or to build additional facilities are cases for using such pumps.
The basic requirement to develop any field is that the wellhead pressure, may be, assisted by secondary recuperation means - water injection, gas lift - is enough to transfer the multiphase effluent through a single, multiphase pipeline. Otherwise, a pump is needed. pump is needed. A line will be mainly characterised by the flow rate (bbl/d), the composition of the fluid i.e the gas fraction GLR [gas to liquid ratio expressed as a percentage by volume (%) or as a ratio (m3/m3) in percentage by volume (%) or as a ratio (m3/m3) in suction condition], the pressure at the pump inlet and the pressure rise needed to enter the first stage of the separator at the end of the line. Secondary data will be PVT parameters such as fluid density, temperature, viscosity, water cut etc...
Considering the above, the wide range of possible combinations must be emphasised, especially for the gas fraction that vary from zero in the case of a pure monophase fluid, to 95% for wells where a gas lift production is in use or if the produced gas has production is in use or if the produced gas has already reached high GOR, 100% being reached if there is severe slugging.
In this work mathematical concepts and fundamentals of a model for well testing evaluation of layered reservoirs are presented. The model, based on that developed by presented. The model, based on that developed by Eligh-Economides et al considers the production to the well as coming from several layers. These layers may be different, regarding lithology and fluid properties. The solutions allow the calculation of the flowing pressure, average pressure, skin and flow rate for each individual layer. This is important for pressure transient analysis since the conventional methods only pressure transient analysis since the conventional methods only provide the global pressure for all the layers. Crossflow is provide the global pressure for all the layers. Crossflow is included in the model. The solutions with model proposed are compared with those given by a numerical simulator in order to check the validity and are in excellent agreement with the numerical simulator. Several field examples are presented to show model applications. Individual layer evaluation can be obtained when testing a well completed in several layers by the model presented Also, production decline monitoring is possible for each layer production decline monitoring is possible for each laye
Shales are usually present in oil and gas reservoirs. Consequently, different zones in the reservoirs are separated by them. The communications between different zones depends on the dimensions and permeabilities of the interbedded shales. These reservoirs are known as "multilayered". When the layers communicate only through the wellbore, we have a commingled system. If there are some degree of communication between two or more layers. we have a crossflow system. Reservoir parameters estimation for individual layers is important for the overall behavior of the reservoir. Using logs, the engineer may infer the layered nature of a reservoir. Lags indicate thickness, porosity and position of all the layer. However, logs can not indicate the extension and sealing made of the different shales. Well test from layered reservoirs may present crossflow between layers. This crossflow is more severe when the layers have different pressure and/or drainage radius. In general it has been observed that conventional well test analysis usually provides pressure only for the total system, being difficult to provides pressure only for the total system, being difficult to distinguish the behavior of a single layer formation from a multilayer reservoir. Multilayer well testing is relatively new to the petroleum industry. It has been used primarily to measure individual layer productivity by measuring downhole flow rates and pressures. productivity by measuring downhole flow rates and pressures. Specifically, a multilayer test (MTL) can yield the permeability and skin values for each layer. The layered reservoir problem has been addressed in many publication according to the literature. The need to develop publication according to the literature. The need to develop technology and software in this area was motivated by two reasons: 1. Conventional pressure transient analysis provides only weighted average permeability for the total producing interval and in many cases the obtained value is subject to discussion; 2. Individual testing of layers is very expensive. On the other hand it can not be done in many cases because of the well completion.
This paper reports ongoing work at Stanford University aiming at better understanding of steam-foam flow behavior with oil present. Steam foam systems were studied in a one dimensional sandpack at residual oil saturation (about 12%). The strength of the in-situ generated foam, indicated by pressure gradients, is significantly affected by injection procedure, slug size, steam quality and non condensible gas (nitrogen). In the range studied, surfactant concentration has a minor effect.
In the presence of residual oil, simultaneous injection of steam and surfactant fails to generate foam in the model, while the same procedure generates a strong foam without oil. When surfactant is injected as a liquid slug ahead of the steam, foam is generated, albeit at a lower strength than when no oil is present. The minimum slug size required to generate foam by present. The minimum slug size required to generate foam by this procedure is about 5% of the pore volume at 1.0% weight concentration. Above this minimum, increases in slug size or in concentration had little effect. The quality of the steam when it is increased, improves the foam strength. One implication of these results is that savings in the amount of surfactant needed for a steam-foam project may be possible by using a large slug of dilute solution. That is, if the slug volume is above some minimum, a dilute concentration may give acceptable foam generation. The reverse procedure, a small slug of high concentration surfactant gives poor foam generation and should probably be avoided in the field. Steam quality effects are consistent with the existing literature.
Foaming additives have been used extensivelly in the field to improve steam injection efficiency (Castanier and Brigham). The purpose of foaming additives is to reduce gravity override and channeling to injected gases and improve gas/oil ratios and sweep efficiency. Despite numerous studies dating back to the 1950s the mechanisms involved during the flow of steam/surfactant/noncondensible gases in an oil reservoir are not fully understood. Modelling of steam foam processes in order to predict production of oil is not yet possible. Optimization of the injection method, amount, concentration and type of surface active agent, relative amounts of steam and noncondensible gas, and steam quality is needed for better application of this enhanced oil recovery technique.
Previous studies have focussed on selecting thermally stable Previous studies have focussed on selecting thermally stable surfactants suitable for use as foaming agents at steam injection conditions. Most of this work was performed without oil or under static conditions. Recent work by Hamida et al. showed that at his experimental conditions it was not possible to generate foam as seen by increased pressure gradients when residual oil saturation to steamflood was present. The procedure that they used was coinjection of steam, nitrogen and procedure that they used was coinjection of steam, nitrogen and surfactant solution. Other authors (Demiral et al.) reported success with the slug injection method although at different conditions. This paper attempts to study steam foam mechanisms at residual oil saturation under dynamic conditions by using an alternating injection technique at the same experimental conditions previously reported by Hamida et al.
The equipment schematic is presented in Figure 1. The main components are injection and production equipment and a cylindrical tube of 2m length and 5cm diameter. The tube is instrumented with 21 thermocouples and divided in 4 sections for pressure gradient measurements.
The pipeline transportation of petroleum fluids can be significantly affected by flocculation, deposition, and plugging of asphaltene, paraffin/wax, and/or diamondoid in the transfer pipelines, tubulars, pumps, and other equipment. The economic implications of the problem of heavy organic depositions in such processes are tremendous.
In this report Mexico's experience with the pipeline plugging due to heavy organic deposition is reviewed and analysed based on the present state of knowledge. The modeling basis of a comprehensive estimation and prediction technique is presented. This technique is based on observed field and laboratory data, statistical mechanical theories, polydisperse polymer solution theory, continuous thermodynamics, electrokinetics and transport phenomena, colloidal solution theory, and the FRACTAL aggregation theory.
A number of problems could arise during the production of petroleum which would drastically increase the production costs. Among these, wax deposition due to a drop in temperature, and asphaltene deposition due to a variety of causes, in the production tubing strings are the prominent problems [l-4]. Therefore, it is of great importance to understand the behavior of heavy organics under various operating conditions. An ultimate goal is to predict whether organic deposition will take place, and to be able to avoid getting to the onset of organic deposition region.
Surveys of field experiences [5-7] indicate that asphaltene deposition problem is one of the major factors that increases the production cost. Thus, being able to prevent such deposition, costs could be lowered appreciably. Mexican oil production and its heavy organic deposition problem was selected for this study because of the availability of an abundance of field and laboratory data about it. Laboratory analysis performed on asphaltenes extracted from samples of the crude oils prone to asphaltene deposition, and from the heavy organic deposits revealed that "Mexican" asphaltenes are very similar (in elemental analysis) to asphaltenes extracted from other sources, as can be seen from Table 1.
The difference would reside in the asphaltene-particle structure and charge which in the case of Mexican crude oils is negative [6,7]. The molecular nature and structure of the asphaltene fraction of crude oils has been the subject of numerous investigations. There are still serious shortcomings in consistency between such studies because of the varied assumptions that have to be made in deriving the molecular formulae.
Franseen, R. (Koninklijke/Shell Exploratie en Produktie Laboratorium) | Vahrenkamp, V.C. (Koninklijke/Shell Exploratie en Produktie Laboratorium) | Van der Graaff, W.J.E. (Koninklijke/Shell Exploratie en Produktie Laboratorium) | Munoz, P.J. (Maraven S.A.)
The Deep Cretaceous carbonate reservoirs of Lake Maracaibo, Venezuela, produce mainly from open fractures. Any improvement in production rate requires optimal access of the wellbore to open fractures. Well tracks with maximum rates of open fracture interception have been calculated using computer models for a structure defined by the Icotea fault in Block IX. Optimum well tracks are towards azimuth 330° with deviation 60° in the West Flank and towards azimuth 030° with deviation 60° in the East Flank.
The open fracture networks consist of incompletely cemented or leached fractures which contain a "channel and island" structure of interconnected porosity. This structure, in combination with the high rock matrix strength and the subparallel strike of the open fractures relative to the E-W trend of the present-day maximum horizontal stress, indicate that it is unlikely that the conductivity of individual fracture will be reduced drastically upon pressure depletion.
The reservoir geological model established for Block IX is consistent with data obtained from the West Flank of Block I, some 40 km to the north and therefore has regional validity.
The Cretaceous carbonate reservoirs in Lake Maracaibo are found beneath the productive Tertiary clastic reservoirs and are referred to as the "Deep Cretaceous", since they lie at a depth in excess of 10,000 ft (approx. 3000 m). Production of hydrocarbons from these Deep Cretaceous carbonate reservoirs is controlled by open fractures that connect the matrix and stylolite porosity with the wellbore1. Any improvement in production rate requires effective access to the open, hydraulically conductive fractures and knowledge of their occurrence with respect to stratigraphic architecture. This will help to: 1) optimize the interception rate of the hydraulically conductive fractures2; 2) predict the pressure sensitivity of the reservoir; and 3) establish the regional validity of the reservoir geological model.
This contribution intends to show how a detailed description of the fracture network, based on core observations and the relationship of open-fracture development relative to the stratigraphic architecture, helps to optimize field development. Rock mechanical tests have been performed to assess the effect of reservoir pressure decline on the hydraulic conductivity of the open fractures.
One important goal in reservoir testing is to determine the reservoir characteristics and the well's productivity. Measurements which can contribute to this goal include well logging, fluid sampling, pressure transient testing, and production profiling. A careful analysis of these measured production profiling. A careful analysis of these measured data will provide adequate information for well evaluation. The purpose of this paper is to show how an integrated reservoir purpose of this paper is to show how an integrated reservoir testing methodology in conjunction with a human engineered procedure on computer enhances the reservoir characterization procedure on computer enhances the reservoir characterization and the evaluation of well performance.
This integrated methodology includes concepts, which have been implemented in recent approaches to reservoir testing and NODAL* analysis. During the design phase of each test, sensitivity studies provide insights that facilitate adequate data gathering for valid interpretation, which can help minimize the cost of the test. During interpretation, the central issue is to identify a realistic reservoir model which reflects the transient pressure and rate behavior. Here significant use is made of the log-log plot of pressure change and its derivative with respect to superposition time. Systematic diagnosis of the log-log plot helps identify different flow regimes, while specialized plots and type curve analysis provide a basis for the estimation of reservoir parameters. Lastly, the computed results can be easily integrated into a NODAL analysis to study and evaluate the well's productivity.
The examples presented illustrate the benefits of a systematic approach to reservoir test design and interpretation, and well productivity evaluation. The computer procedure also has the advantage of incorporating input data from several sources, including openhole logs and the Repeat Formation Tester (RFT*) tool.
During well testing operations, pressure, flow-rate and temperature transients are generally recorded using testing devices and the Production Logging tools (PLT). With the tools in a stationary position in the wellbore, the transient behavior of the reservoir is observed by varying the surface flow-rates. On the other hand, under stabilized and shut-in conditions, production logging measurements are conducted versus depth. Interpretation of these different types of data help the engineer to quantify the well/reservoir model, and to diagnose well performance.
The methodology outlined in this paper integrates a test design approach, currently acceptable acquisition procedures, interpretation techniques and NODAL analysis. procedures, interpretation techniques and NODAL analysis. For both vertical and horizontal wells, the interpretation procedures include the log-log presentation, the use of procedures include the log-log presentation, the use of pressure derivative for identifying dominant flow regimes, pressure derivative for identifying dominant flow regimes, convolution type curves and the convolution derivative when downhole flow-rates are available, and an automatic parameter estimation. Lastly, the integrated reservoir parameter estimation. Lastly, the integrated reservoir engineered methodology shows how interpreted results can be easily utilized to evaluate the performance of a well via NODAL analysis./
Several examples are presented to illustrate how the integrated methodology of test design, data acquisition, interpretation and well evaluation was encapsulated into a human engineered computer environment. The first example illustrates the advantage of downhole shut-in during a buildup test. The second example illustrates how different aspects of the interpretation procedure aid in evaluating late time effects. The next example demonstrates the transient response of a horizontal well test, the flow regimes that occured and the analysis thereof. Lastly, the evaluation of the horizontal well performance is compared with that of a vertical well using the NODAL analysis.
Conduction calorimetry technique makes possible to follow the kinetic of cement hydration through the rate of heat evolution. This work presents experimental results using the following types of cement: oilwell Class G, Class A, Pozzolanic and Blast Furnace. The effects of cement aeration (carbonating and/or humidifying), temperature and use of additives on the hydration rate of Class G cement slurries were analyzed. In addition, the characterization of the different cement types was performed. The tests were run in a JAF Conduction performed. The tests were run in a JAF Conduction Calorimeter which working principles were described in the work. A computer program was developed to calculate the hut liberation rate and the cumulative hut of hydration generated along the process.
The conduction calorimetry is used to correlate the heat of cement hydration and the changes in physical properties of Class G cement. These changes are caused properties of Class G cement. These changes are caused by cement aeration which occurs due to cement transfers: and storage from the manufacturer to the oilwell.
The primary intentions of this work are to discuss the usefulness of this technique to the oil industry and to suggest other possible applications.
THE JAF CONDUCTION CALORIMETER
In this section, the basic principles behind the conduction calorimetry technique are discussed. The JAF calorimeter cell (figure 1) consists of a sample can (figure 2) and of a flanged acrylic cylinder, having a set of thermopile sensors, which are in fact thermoelectric (Peltier) hut pump modules constructed of N and P elements of high grade bismuth telluride (1). The hut produced during the cement hydration flows by produced during the cement hydration flows by conduction through the oil causing a temperature increase which are detected by the sensor: and registered on the plotter and on the digital reading in the Interface Module. Since the convection and radiation losses are negligible the apparatus is called conduction calorimeter.
The heat of rate liberation is calculated with Tian-Calvet equation, resulting from an energy balance:
The thermal electromotive force (E), developed by the thermopile sensors during the process is directly proportional to the temperature difference between the proportional to the temperature difference between the sample (T) and water bath (T):
The constants k and k are not know previously. Moreover, these constants depend on the sample and can geometry and on their thermal properties and densities. It follows that such constants must be determined specifically for each sample can system being tested.
The preparation of the slurry used in the conduction calorimetric test is performed in accordance with API procedures tests. After removing the cup from the Waring procedures tests. After removing the cup from the Waring Blendor, the core sample is stirred with a spatula by approximately 30 seconds. After that, part of the sample is placed in a 250 ml becker, and the slurry is stirred by a glass stick during 30 seconds more.
As part of the evaluation of a deep sandstone interval in eastern Venezuela, hydraulic fracture employing high strength proppant was used to determine the formation's flow potential. The interval fractured is at a depth of more than 16,000 feet and has a bottom hole static temperature of 288deg.F, conditions which require a good knowledge of reservoir and fluid properties. The design of this operation incorporated properties. The design of this operation incorporated a number of data gathering steps to ensure that fracture geometry, in-situ stresses, and fracturing fluid behaviour were understood prior to executing the treatment.
Mechanical properties logs indicated that two separate fractures would develop at different pressures, the upper fracture initiating some 600 psi pressures, the upper fracture initiating some 600 psi above the initiation pressure of the lower interval. The first pump-in test performed on the well consisted of a ball-out employing 600 balls transported in a linear gel frac fluid. Following this a temperature log was run across both intervals. The deflections indicated that two fractures had indeed been generated, both fractures coinciding well with the height predicted by the sonic log, although the upper fracture had grown some 50 feet higher than the prediction.
High temperature crosslinked water based gel was pumped as a minifrac without diverter (no ball pumped as a minifrac without diverter (no ball sealers) at a rate of 18 bpm as the second downhole pumping test. This part of the operation was pumping test. This part of the operation was designed to determine actual fracture geometry, fracture gradient, and particularly the leakoff coefficient of this fluid.
The data on fracture height, fracture gradient, post-fracture pressure decline and fluid loss obtained post-fracture pressure decline and fluid loss obtained from the logs and pump-in tests were incorporated into the design of the main fracture and the operation was successfully completed. The paper describes each of the pumping tests and shows how the test results and log derived data were used to verify the eventual design of the propped fracture.
The success of a hydraulic fracture is strongly dependent on an understanding of orientation and overall geometry of the created fracture, which in turn are primarily functions of the orientation, magnitude and distribution of in situ stresses in the reservoir an adjacent strata.