Abstract This paper presents a PC based alternative procedure for determining the water saturations within the hot water zone of a thermal project for use in analytical oil recovery calculations. Conventional analytical calculation of oil recovery under steam and hot water injection requires the tracking of the movements of the saturations advancing within the variable-temperature hot-water zone. This involves an adaptation of the Buckley-Leverett theory to this variable temperature zone after dividing it into a number of constant temperature or isothermal zones. If the number exceeds two, the calculation can become very tedious unless done with a computer program. FORTRAN programs have generally been used but they are not as intuitive or easy to use as modern PC based programming tools such as MathematicaTM. In this paper. MathematicaTM was used because it is relatively easy to program. easy to use, and is fast and robust. Unfortunately. depending on the number of isothermal zones, the time step size and the time at which the recovery calculation is desired, the tracking of saturations can tax the capabilities of even the most modern PCs. Therefore, an alternative method is introduced that eliminates the need for saturation tracking. This method calculates the instantaneous saturations within each isothermal zone at the time of interest. Oil recovery results by this method were found to be comparable to those by the saturation tracking method with considerable saving in computation time. Two examples are presented to demonstrate the utility of the method. Introduction For one-dimensional models, the analytical calculation of steamflood oil recovery with time requires calculating the following: the position of the steam front and its rate of advance or velocity. the temperature profile in the hot water zone and its rate of advance, the fluid saturation profile and its rate of advance. The location of the steam front at any time can be calculated by the equations of Mandl and Volek who showed that before a critical time tc, the steam zone can be described by the equations of Marx and Langenheim. Beyond tc, Mandl and Volek present an approximate equation to calculate the steam front location with time. This equation was later improved by Prats and Vogiatzis and communicated to Myhill and Stegemeier who presented the solution in graphical form. The temperature profile in the hot water zone can be calculated using the equation by Lauwerier for both hot waterflood and steamflood. However, for a steamflood, a simpler approximation assumes a linear temperature drop from the steam temperature to the cold reservoir temperature. The saturation profile in the steam and hot water zones can be calculated using the Buckley-Leverett theory provided that the non-isothermal hot water zone is first divided into an appropriate number of isothermal zones. Several procedures are available to calculate the saturation profile with time. This paper is concerned with the method whereby characteristic saturations are picked and their motions tracked. The tracking process is very time consuming and can easily exceed the capabilities of PCs' when small time steps are used together with many isothermal zones at long times. An alternative to tracking is used whereby instantaneous saturations in each temperature zone are calculated at any given time. The saturations to be tracked can be picked in different ways. Farouq Ali determines the flood front saturation at the cold reservoir temperature (just as in a waterflood) and arbitrarily picks saturation values greater than this value ending at the maximum water saturation. This has the advantage of giving greater saturation definition where needed. However, it has the potential of using more saturations than is needed leading higher computation time on personal computers. Willman et. al. suggest a somewhat narrower range of saturations by finding as the starting point, that saturation in the cold zone that have the same velocity as the cold temperature at tc. Prats extends this further by finding a particular saturation for each isothermal zone referred as characteristic saturations. These are calculated as the saturation in each zone that has the same velocity as the cold temperature at tc. This way, there are only as many characteristic saturations to be tracked as there are isothermal zones as opposed to tracking an arbitrarily large number of saturations. Even with these improvements in the choice of saturations, the tracking process of calculating their locations at each time step is time consuming even for personal computers. P. 83

NodalBc: A Unique Program for Optimum Production of Sucker Rod Pumping Oil Wells Abstract Sucker rod pumping is the oldest artificial lift method used in production of oil wells. This method is usually designed and analyzed with the assistance of programs based on simplified models, such as the API RP 11L, or with a wave equation simulator. In many cases, however, a technique that permits consideration of the inflow and outflow performance of the well is a more efficient way to design and analyze an oil well. This technique, referred to as Total Analysis or System Analysis (SA), has been successfully applied to flowing and artificially lifted wells. This paper presents a computer program developed by Intevep S. A., called NodalB, created to perform the SA of sucker rod pumping wells. NodalB 's unique characteristics make it a competitive program in today's market of design and optimization software. This program allows consideration of the inflow performance of the reservoir depending on the drilling scheme, either vertical or horizontal, and determination of the sensitivity of the rod pumping system to its most relevant parameters, such as: pumping speed and stroke, pump diameter, gas oil ratio, gas separation efficiency, tubing diameter, pump depth, rod string design, oil viscosity, injected diluent (diesel, kerosene, etc.) ratio, water cut, reservoir pressure, productivity index, and different production schemes (i.e. production through the tubing or through the casing-tubing annulus). These features show the power of this program as a predictive tool. NodalB is specially applicable to design, optimization and troubleshooting of heavy and viscous oil wells, taking into account the viscosity effects in the rod pumping system. Introduction The current total oil production of Venezuela is approximately three (3) millions of barrels per day. One million of it is produced from almost 7000 sucker rod pumping wells, and consists, in its majority, of heavy oil and bitumen, which typical downhole viscosity ranges from 500 cp up to 7000 cp. The corresponding API gravity ranges between 8 and 12. Due to the high viscosity of the oil, cyclic steam injection has become a typical production scheme for those wells in which static reservoir pressure is below 1000 psi (up to 2500 ft of well depth). Also, other alternative production schemes have been developed to reduce 1) the drag forces between the rod string and the fluid and 2) the pressure drops in the tubing string and flowline. Such schemes involve the injection of a low viscosity fluid (diesel, kerosene) in a continuos basis into the wellbore and flowline. Furthermore, the oil being produced usually flows through the tubing-casing annulus instead of the tubing. Recently, a new completion scheme, called the Downhole Stuffing Box Production Scheme, has been successfully tested in the field, revealing several advantages over the other heavy oil production schemes currently applied in Venezuela. P. 33

Abstract A comprehensive formation evaluation method is presented which integrates log, core and well test data on a reservoir using regression analysis. The method employs a capillary pressure curve model correlating four reservoir variables: porosity, water saturation, permeability and capillary pressure. Porosity and saturation are estimated by conventional log analysis, permeability is obtained from empirical correlations with logs (usually porosity) and capillary pressure is directly related to height above the water level. Depth profiles of the four variables are adjusted by regression analysis constrained by the capillary pressure curve model. Results of the regression include An estimate of the water level, Improved profiles of the four variables which are consistent with both log analysis and capillary pressure theory, Adjusted log analysis parameters, capillary pressure level, curves (drainage) capillary pressure Complete synthetic curves for each depth Relative permeability generated from the curves, and Estimated effective permeability to hydrocarbons and to water opposite the wellbore. These results are then integrated with well test data by comparing effective permeabilities from 6. above with the test results. If there is a mismatch, it may be necessary to rerun the regression in order to honor all of the datasets. The method is illustrated with log, core and test data from an offshore Gulf of Mexico well. Introduction There are three classic methods for estimating water saturation in a reservoir: log analysis, capillary pressure curves and oil-base coring. By far, the most common method in use today is evaluation by log analysis alone. P. 213^

Abstract Three-dimensional fluid flow through porous media problem,; have assumed increased significance and complexity in recent years due to a growing interest, in horizontal well applications. Various predictive tools based on analytical and numerical methods have emerged to address such problems. They have various drawbacks. A versatile boundary-element algorithm was developed in this work to provide a viable alternative. The algorithm is capable of solving transient and steady-state problems in isotropic or anisotropic reservoirs and handling finite reservoirs and finite-radius well(s) of arbitrary geometry. It also allows the prescription of arbitrary combinations of the two common types of boundary conditions (Dirichlet and Neumann). It can thus be used to solve a wide variety of three-dimensional horizontal well problems in relatively complex situations. A major advantage is gained due to the need to discretize only the domain boundary and not the domain itself, thus effectively reducing the dimensionality of the problem by one and eliminating grid orientation effects and numerical dispersion. The reduction in problem dimensions allows three-dimensional problems to be modeled using two-dimensional grids. This paper provides a summary description of the main features of the algorithm. The application of the algorithm is demonstrated using two examples involving horizontal wells. Introduction Potential benefits in increased productivity and recovery efficiency coupled with advances in drilling and completion technology have roused considerable interest in horizontal well applications. Since horizontal wells are characteristically less symmetric (compared to vertical ones), and horizontal wells typically penetrate the reservoir more and in paths that are more tortuous, the flow field and pressure distribution created by such wells can intuitively be expected to be three-dimensional in nature and thus be more susceptible to irregularities in geometry and other system properties in three dimensions. That the horizontal well problem is three-dimensional in nature has been explicitly or implicitly expressed by various authors and a good example of real-life irregularities is provided by Joshi, who gave a typical horizontal well profile (Figure 1) and pointed out that horizontal wellbores are rarely horizontal but, rather snake-like with many variations in the vertical plane. Thus, the emergence of horizontal well applications has had two significant implications. P. 91^

Abstract The type curves developed by Fetkovich are widely used tools for decline curve analysis. Fetkovich combined the analytical constant pressure infinite (early transient period for finite systems) solutions with the "empirical" exponential, hyperbolic, and harmonic decline curve equations developed by Arps. There are software applications that have computerized this analysis technique; however, attempts to automate the matching process have been hindered by two problems: a way to identify the transition from one flow period to the next, and a way to minimize the adverse impact of outliers, data considered to be bad observations. This paper presents methods to handle these issues. Recognizing the point of transition between flow regimes is sometimes difficult, even when the matching process is done manually. If, however, the curves are regenerated using the proper derivatives of the production, the software can locate the transition simply by looking for a minimum. The process of estimating reservoir parameters is simplified as well. To handle outliers, this paper proposes a robust algorithm consisting of a nonlinear parameter estimation using Least Absolute Value (LAV) as the criterion for minimization. One advantage of this approach is that it provides a smooth transition between full acceptance and total rejection of a given observation. Outliers are systematically assigned less weight in the objection function. This method also avoids the computation of a second derivation of the model function. Most importantly, tests have shown that when outliers are introduced into the observations, the LAV algorithm still produces accurate parameter estimates. The above techniques have been implemented in software that operates in the MS-Windows TM environment. With its graphical user interface (GUI), this environment allows for clear visual displays of the matching process and the results. Moreover, it provides the user with many more options than the traditional, nongraphical operating systems. Because programming under MS-Windows is still relatively new to the industry, this paper also discusses some aspects of Windows programming, especially as it relates to petroleum applications. Introduction Decline curve analyses are used to interpret rate-time data, to estimate properties, and to predict future performance of the reservoir system. Fetkovich's type curves combined the analytical constant pressure infinite (early transient period for finite systems) solutions with the empirical exponential, hyperbolic, and harmonic decline curve equations developed by Arps. Fetkovich et al. demonstrated the application of type curves through numerous case histories. Various author have considered production rate decline analyses given a constant pressure drawdown. Ehlig-Economides and Ramey and Uraiet and Raghavan published general studies about transient rate decline curve analysis. The analytical solutions presented by these authors form the transient portion of Fetkovich's type curves. The studies pointed to the reciprocity relationship between wellbore flowing rate and pressure. This relationship was used to draw some analogies between constant rate and constant pressure testing. Ehlig-Economides and Ramey described some practical methods for transient rate analysis of wells produced at constant pressure. Reservoir permeability, porosity, and well skin factor can be determined from the type-curve match, and exponential rate decline can be analyzed to determine the drainage area of a well and the shape factor. Uraiet and Raghavan also discussed methods for determining the formation flow capacity (kh) and skin factor (S) through type-curve matching. In addition, Turki et al. presented the parametric behavior of a constant pressure well in the center of a two-composite reservoir with wellbore skin and rate derivative type curve. Their results complement the rate decline curves of Fetkovich. There are software applications available that have computerized Fetkovich's type curves. To our knowledge, most of these applications are based on manual type-curve matching. P. 229^

Abstract A new drilling model has been developed for the interpretation of downhole drilling conditions using surface and downhole drilling data. This model is based on the analysis of the interaction of the drill bit (PDC or impact) and the formation under in situ pressure conditions. It combines downhole and surface drilling data with MWD formation measurements to provide detailed drilling analyses for real-time applications. An expression for Drilling Response developed from the model provides information on lithologic changes and allows interpretation of downhole drilling conditions. Drilling Response is a function of lithology, pore pressure, mud pressure and bit properties and is computed from the following drilling parameters: rate of penetration, bit revolutions, and downhole weight and torque on bit. Techniques and expressions allowing computation of formation porosity and pore pressure on a foot by foot basis have also been developed based on the Drilling Response. Another key feature of the model is the identification of inefficient drilling conditions. Inefficient drilling conditions are expressed in the form of Drilling Alerts for real-time applications, allowing the operator to take the necessary action to rectify the problem at the wellsite. Drilling Alerts also allow the operator to continuously monitor the effects of any remedial actions. Bit condition is evaluated using Drilling Response and torque analyser. This allows the operator to assess bearing and tooth wear conditions for impact bits and cutter wear for PDC bits. Field Examples from sedimentary basins illustrate the usefulness of the model and its applications at the wellsite and in post-drilling analysis. Introduction The advent of Measurement-While-Drilling (MWD) has made it possible to obtain measurements of drilling parameters very near the bit, and these measurements can provide significant information about the formation being drilled and the condition of the bit during drilling. Among the measurements that can now be made by an MWD sub placed near the bit are weight on bit (WOB), torque on bit (TOB), formation gamma ray, and directional measurements. Rotational speed (RPM) and rate of penetration (ROP) are calculated from surface measurements. Due to friction along the drill string, weight applied to the formation at the bit is generally lower than weight measured at the surface. For the same reason, surface measured torque may not bear any resemblance to actual torque at the bit. Because the torque measurements are made near the bit, they more accurately represent the response of the bit to formation interaction. ROP is also a reflection of bit/rock interaction; consequently, for a fixed WOB, RPM, and mud density, variations in these two parameters should provide a direct indication of formation changes. P. 275^

Abstract This paper describes the use of residual analysis in production rate decline curve fittings and performance predictions. Residual analysis, in performance predictions. Residual analysis, in statistical term, is an easy and quick check of any assumptions being violated under certain hypothesis test, leaning an attempt to discern regression drambacks or pitfalls if any. In production rate decline analysis, residual analysis technique is used to demonstrate the validity of decline equations or models for given rate-time data, and the reliability of the estimated parameters, such as qi, n, Di, and then that of future parameters, such as qi, n, Di, and then that of future performance predictions. The emphasis of the study is on performance predictions. The emphasis of the study is on the detecting analysis of Arps' decline equations, with other forms of empirical expressions included. For the hyperbolic equation which is slightly lore difficult to solve, eight typical numerical solutions are analyzed. These solutions do not often reach good agreement with each other for field data. Graphical residual analysis shows which solution fits the decline trend best, and which thereafter gives reliable performance and reserve predictions. A microcomputer-aided commercial software predictions. A microcomputer-aided commercial software is developed and has found applications in China's oilfields. Introduction Decline curves are one of the lost extensively used forms of data analysis employed in the evaluation of oil properties and the prediction of future behavior of a properties and the prediction of future behavior of a well or lease. The general problems in decline curve analysis involve selecting a right equation to watch declining production rate-time data or increasing cumulative production data and then making extrapolations for future performance predictions. Over the past decades, many investigators studied the different aspects of the topic such as the ability of the range of hyperbolic exponent, the solution of hyperbolic equation, and so on. In practical application of decline curve analysis, for a set of given production data, it is an usual step for a reservoir engineer to pick up one what is said "best-regressed" Mathematical model or decline equation to fit production trends or to use a French curve or a series of constant-percentage decline to process the data. It is well known that the parameter estimation results obtained from any method based on statistical regression treatment, especially those from the least-squares will always need backward testings and/or secondary detectings. The purpose of these tests and detections is to discern any problems possibly occurring in the data treatment. In the concrete, the problems may include: how properly the model is chosen, to what extent the estimated model constants are accurate and reliable, if there are any calculation mistakes, and if there exist sole abnormal data points (or outliers) which need extra processing, and the like. Although in the past studies, sole people had limited work with the topic in the aspects of statistical tests of assumptions, concerning the determination of confidence intervals, Durbin-Watson automation and heteroscedasticity, there is no serious consideration of secondary detections for decline curve analysis, and this is really necessary according to experience. This paper presents a supplementary detecting technique with graphical residual analysis supported by PC microcomputers. Through such detectings. sole PC microcomputers. Through such detectings. sole possible pitfalls in the analysis light be found which possible pitfalls in the analysis light be found which consequently enables the future Production performance and reserve predictions more reliable and accurate. P. 233

Abstract A kick simulator has been developed for use on a work station with Unix operating system and X Window System installed. Advanced mathematical models are used to calculate the development of gas kicks and oil kicks in water based and oil based drilling muds. In the preprocessor the user sets up the input data in an environment with help functions and default values. When the simulator is running, the user can observe a number of selected surface and downhole kick parameters vs. time and space plotted graphically on the screen. In the postprocessor the user can display different physical parameters vs. time and space on the screen and make hard copies of them. Introduction A kick while drilling represents a potentially very dangerous situation. It is of primary importance to detect the kick as early as possible and to circulate the kick to the surface in a safe way without creating problems like lost circulation or secondary kicks. Todays drilling in deeper prospects, with frequently small margins between pore pressure and fracture pressure, puts increased demands on the pressure control while circulating out a kick. The traditional kick models, assuming a very simplified model for the transport of the kick fluid (gas, oil), are too inaccurate to be used for planning and design purposes. A summary of published kick models have been purposes. A summary of published kick models have been given by Element et al. The Rogaland Research Kick Simulator has been developed in order to make a user friendly, verified, planning and design tool for drilling engineers. Results from sensitivity studies of gas kicks in oil based muds, and comparisons between gas kicks in water based and oil based muds have been published. published. The model is validated with data from gas kick experiments in a 60 deg. inclined, 2000 m long research well. These experiments included both water based and oil based muds and were extensively instrumented, giving excellent data for model verification purposes. In order to improve the user friendliness of the model, a new user interface on a Unix/X Window/Motif basis has been developed. This will greatly improve the usefulness of the model, and make it easy accessible to drilling engineers for design and planning purposes. In addition the new user interface will greatly enhance the model capabilities as a training tool, since the user can follow the impact of his control measures interactively on the screen plots. This paper briefly presents the model basis, and then plots. This paper briefly presents the model basis, and then details the user interface capabilities. MATHEMATICAL MODEL The fluid flow in the well is considered a combined single and two phase flow problem. We restrict ourselves to one spatial dimension i.e. position along flow line s. P. 249

Abstract A mathematical model of hydrodynamic processes i intermittent gaslift is described. The many facets of fluid flow accrue from continuous changes of the hydrodynamic and thermodynamic conditions, the mathematical solution of unsteady fluid flow was simplified by introduction of certain assumptions and boundary conditions. These synthetic. analytical solutions were used to develop the "POVLIFT" program for designing and optimizing intermittent program for designing and optimizing intermittent gaslift. Introduction Intermittent gaslift is an artificial, oil production method often applied to low productivity production method often applied to low productivity wells. The technique includes controlled injection of compressed gas into the tubing-string. through the gaslift valve, and relatively rapid lifting of portions of the fluid accumulated in the wellbore. portions of the fluid accumulated in the wellbore. The high velocity gas bubles penetrate into or pass through the fluid slug constantly reduction its initial length due to fallback to the bottom of the well as condensation film. The unsteady nature of fluid flow in the reservoir and borehole and the rapid changes of hydrodynamic and thermodynamic have hither to hampered the development of a mathematical model which would simulate the complex dynamic processes and describe the behaviour of all relevant processes and describe the behaviour of all relevant parameters of such a dynamic system as the combined parameters of such a dynamic system as the combined reservoir and borehole conditions prevailing in oil production by intermittent gaslift. The hydrodynamic production by intermittent gaslift. The hydrodynamic model integrates all of significant variables of two essential subsystems: the reservoir - or the fluid flow in the reservoir, and the well - or behaviour of the fluids in the wellbore. The establishment of optimum production conditions involves adjustment of the essential parameters of both subsystems (the physical properties of the fluids, the hydrodynamic physical properties of the fluids, the hydrodynamic properties of fluid flow in the reservoir, bottom properties of fluid flow in the reservoir, bottom hole completion, the respective velocity of liquid and gas…), in such a way that will provide maximum utilization of the injected power due to the favourable relation of the physical phenomena and the phase effects during the lifting of the liquid slug. The design of the mathematical model is oriented primarily towards reconciling and consolidating the ostensibly different subsystems (the reservoir - the well- the operating parameters and their interaction) using well prepared and selected data and various, interchangeable program modules. 1. DESCRIPTION OF THE MATHEMATICAL MODEL FLUID FLOW THROUGH THE RESERVOIR The proposed mathematical model of fluid flow in the reservoir includes various methods generally used for defining one or two phase inflow into the wellbore. These methods are described in the references (Ref.1.2.3). Fig. 1 the general flow chart used to develop the specific algorithms and the related programs. Reservoir pressure decline depends on several factors. the most important being the physical properties of rocks and fluids, flow regime, oil properties of rocks and fluids, flow regime, oil saturation, formation damage and general state of the borehole area, displacement mechanism etc. Once the porosity and permeability data, indispensable for estimation of in-situ oil, are interpreted from laboratory test data or geophysical well logs usually early in the life of the reservoir, however the laboratory analyses of the fluid characteristics or PVT relations are often neglected so only the PVT relations are often neglected so only the application of adequate mathematical models (mostly empirical correlations derived by regression on the experimental data) can provide the required parameters. No analysis of fluid inflow performance parameters. No analysis of fluid inflow performance can be made- as indicated in Fig. 1 - unless PVT data tested or calculated, are available. Fig. 2 shows the flow chart of evaluation of the physical properties of fluids. The mathematical description of these methods is available in the references (Ref. 4.5). P. 69

Abstract Permeability from Logs utilizes a relatively new Permeability from Logs utilizes a relatively new type of interactive computer instruction that falls in the category of hypermedia. This tutorial includes hypertext, voice, music, graphics, animation, and interactive computation. The idea is to allow the user to select only that detailed information he or she needs — plus incorporate an element of entertainment while learning. The program can be run on the Macintosh using Apple Computer's Hyper Card, or on IBM-compatibles using Spinnaker Software's PLUS operating under Microsoft's Windows 3.0. For the Macintosh version, Hyper Card's object-oriented programming language Hyper Talk was employed. programming language Hyper Talk was employed. The PLUS programming language PPL was used for the DOS version. Introduction Conventional ways of transferring information from research and development laboratories to field-office personnel are by way of internal reports and personnel are by way of internal reports and in-house seminars. These formats work reasonably well when the audience has roughly the same experience level and academic training. The difficulty arises when disciplines are varied or the audience is comprised of seasoned professionals along with relative newcomers in a given area. In this situation, the material is often either too detailed or lacks supporting evidence, depending on the individual's background. This problem of audience diversity is a serious one, even for the most accomplished writer or lecturer. Use of hypertext with a computer-based tutorial is an excellent tool for addressing a varied audience. Hypertext is the linkage of text material in an associative, nonsequential manner that allows users to access information to a level of detail that is needed. This is particularly helpful when the subject involves more than one discipline, where users may require definitions and elaboration in some areas, but not in others. If we now extend the idea of hypertext to include pathways that contain interactive graphics and interactive computation, coupled with animation and sound, we have what is referred to as hypermedia. Here, the user is intimately involved with the program, particularly concerning control over what data is accessed, and in what order. Also, depending on the imagination of the program designer, the user can be entertained in the process. This can make for enjoyable learning and not be just the tedious process of reading information from a computer screen. P. 89

Abstract Basin modelling is modelling of physical and chemical processes which cause the generation of oil and gas in sedimentary basins. Basin modelling is a valuable tool in the search for new reservoirs. Basin modelling has traditionally been run on minicomputers or mainframes due to the demand for high processing power. With the increasing processing capability of todays personal computers and workstations, the road is open for moving the basin modelling to where it belongs; to the geologist doing the basin modelling. The 2D basin modelling system presented here has emphasized on standardized software, using UNIX as operating system and X Windows for interactive graphics. This makes the basin modelling software fit into the open environment with systems ranging from personal computers to mainframes, thus making basin modelling tools available for the geologist. The system is modular, i.e. each geological process is represented as independent modules. The user has full control over the modelling process, and can modify parameters and monitor the result between each step of the modelling process. A geohistory module reconstructs the sedimentary basin through time, taking into account the progressive compaction of the sediments. The fault blocks can be moved by a fault balance module to balance the profile. The subsidence resulting from the weight of the sediments is computed by the isostatic module, which, together with a tectonic module provides an estimate of the palaeo heat flow in the basin. The palaeo heat flow is an important input parameter for the thermal module, which computes the thermal history of the basin. The kinetic module is used to compute the timing and amount of petroleum generation.

SPE Member Abstract The major cause of the increasingly higher software development costs resides in the lack of reuse of existing technology in new applications combined with the costs of maintaining existing codes. Related to these considerations, Computer Aided Software Engineering (CASE) was introduced in the software industry as a methodology aimed at improving the software productivity. Some of these concepts have been used in the development of a modular software architecture to simulate completion operations, such as cementing, gravel packing, and coiled tubing. The open architecture was designed to host, at a minimal cost, incremental developments in the physics of the underlying processes as well as major modifications. The processes as well as major modifications. The methodology of the development and the choice of the software technology are discussed: this includes the language specifications, the steps in documenting the project and the actual implementation. Compromises between" quick prototyping" and maintenance considerations/code performance have often to be made. The criteria for selecting a given strategy are discussed in terms of software metric. Introduction Increased computational speeds have allowed an improvement in the modeling of complex phenomena arising during completion and stimulation phenomena arising during completion and stimulation operations. This includes more sophisticated rheological descriptions of the fluids used during those operations, and increased degrees of refinement in the description of the formation and the wellbore. The progress in computer equipment has affected firstly office systems, with the advent of mini supercomputers, or architectures involving clusters of engineering workstations generally sharing file servers and eventually connected to a high speed (vector/parallel) central unit. Then, field equipments with sophisticated monitoring capabilities haven emerged, generally in ruggedised configurations. The trend is presently to have fault tolerant hardware, self-diagnosing software, control systems, and realtime filtering/identification processes. In the oil industry, as in all other industries, the cost of software development has become the major component in the development of computerized systems, stressing the importance of developing internal procedures to specify, develop and maintain complex software. P. 221

SPE Member Abstract For many difficult reservoir simulation problems, the fully implicit black-oil models are the only efficient and viable means to the solution. Because of their stringent demands on computer's power. The fully implicit simulators are generally run on mini, main or super computers. However, the current trend of 32-bit microprocessors being used in the main-stream Personal Computers (PC) opens a new horizon for reservoir simulation applications. This paper examines the feasibility of a fully implicit black-oil simulator run on these PC's. The simulator uses a general black-oil formulation and can solve 3-dimensional, 3-phase problems with multi-layered wells. problems with multi-layered wells. Two aspects which affect the applicability of person a computer environment in reservoir simulation are addressed. First, various considerations for models's efficiency in program design and coding are discussed PC systems are program design and coding are discussed PC systems are compared. The calculation examples show that the fully implicit black-oil model presented can be run efficiently on the current 1386 type of PC's for problems up to several thousands grid blocks. For putting the advances of personal computer environment in perspective, some runs personal computer environment in perspective, some runs made on the new IBM RISC/6000 320 desktop are also presented. presented. BACKGROUND Even though many black-oil models use IMPES or sequential formulation for the sake of computational efficiency, the fully implicit model is still the only efficient and viable method to solve many difficult reservoir simulation problems. Because of their stringent demands for the problems. Because of their stringent demands for the computer's speed and memory, the fully implicit simulators are usually run on the mini, mainframe or super computers. To be able to run a program on personal computers is always tempting. Because they are easy to access and use, and economic. There also exists a large numbers of utility programs for PC's which are of great help in preparing programs for PC's which are of great help in preparing and reporting the work being done. One can use pc either to run an entire simulation or to prototype the input dataset for a complex field simulation to be run on a large computer. For a long time (under computer industrial's standard), few reservoir simulation applications have been run in the personal computer environment. The reason is simple personal computer environment. The reason is simple they are not powerful enough to meet requirements of intensive computations and large amount of data storage associated with the reservoir simulation calculations. The 16-bit (98 bit with earlier PC's) word length of 80286 processor is inherently inefficient to the floating point (single or processor is inherently inefficient to the floating point (single or double precision) representation. The processor's speed is low and the storage is limited. P. P.

Abstract PRESEM is a computer aided field development training PRESEM is a computer aided field development training system used in the Petroleum Engineering postgraduate and final year undergraduate courses at Imperial College. PRESEM provides an integration of many of the activities PRESEM provides an integration of many of the activities which go into field development from drilling to the design of surface facilities, economic forecasting of pre-performance and application to regulatory authorities. pre-performance and application to regulatory authorities. The core of the PRESEM system consists of a data base which is generated by the system supervisor through a menu driven suite of programs simulating the properties of a given or hypothetical oil field. The data base contains the geological description of the reservoir, rock and fluid properties, presented as mud and electrical logs, RFT and properties, presented as mud and electrical logs, RFT and production test data, core and PVT analyses. These vary in production test data, core and PVT analyses. These vary in accordance with the depth and location of the exploration wells chosen by the user with reference to the formation top surface map. PRESEM also contains many auxiliary programs including PRESEM also contains many auxiliary programs including Schlumberger type log plotting, log-interpretation, pressure analysis, phase package for separator design, pressure analysis, phase package for separator design, reserves estimates by Monte-Carlo analysis and economic feasibility analysis. The paper describes the basic features of PRESEM and presents a case study. presents a case study Introduction Computer aided teaching systems have been used in the Petroleum Engineering MSc and undergraduate courses at Petroleum Engineering MSc and undergraduate courses at Imperial College since 1977. The MSc is a one year 'conversion' course which the PRESEM project plays an important part, allowing students to apply knowledge gained in a wide range of theoretical courses. PRESEM simulates a field exploration and development project in which students are assigned roles as members of an operating company. The operating company sets out to explore a licensed block in UKCS (United Kingdom Continental Shelf) and reservoir properties represent those of typical North Sea oil fields. However, the system is entirely flexible and contains a menu driven Set Up module which allows the 'supervisor' to set up the properties of many different types of fields. PRESEM properties of many different types of fields. PRESEM consists of three principal modules. The Set Up Module, used by the supervisor, rests on a number of equations which are used to generate fieldwide variations in reservoir geometry, geology and petrophysics. The equations are given sufficient flexibility to match a range of reservoir characteristics. The Exploration Simulation Module allows each operating company to 'drill' a number of wells into the reservoir model. The maximum number of wells which can be drilled is limited at the discretion of the supervisor from the Set Up Module. PRESEM now presents the students with a mudlog, cores, electrical well logs, production tests and PVT data specific to the coordinates of the field in which the well is drilled. The Tools Module consists of a range of special applications software, such as those for pressure analysis, log interpretation, regression based EOS, etc. These, along with other computer systems and analytical techniques, taught in the lecture courses, allow the operating company to analyse the well test data to evaluate potential recoverable hydrocarbon reserves, and hence, to set out and to make an economic assessment of the development plan. At Imperial College the project culminates with a plan. At Imperial College the project culminates with a public hearing in which each operating company, typically public hearing in which each operating company, typically consisting of five students presents their case and is questioned by representatives from industry. P. 199

Abstract An interactive data input program has been developed in the C language to aid creation of data files for five enhanced oil recovery predictive models. The program is designed to run on an IBM PC or compatible microcomputer with transfer of the files to a mainframe for use. It contains several features to assist the end user and speed up data entry including menus, windowing, context-sensitive help, choice lists for variables, and data validity checks. Introduction Simulation and modeling are powerful tools for the petroleum engineer. They can provide insight to reservoir behavior, confirm beliefs or bring new questions to light. Results from a computer simulation depend on a number of factors including model options, accuracy of data entered, and the integrity of the data file. Many reservoir and enhanced oil recovery (EOR) models use data from disk files on the mainframe computer on which the model is running. This is the case with five enhanced oil recovery models, named below, available from the Department of Energy (DOE). The models were developed by Scientific Software - Intercomp for the National Petroleum Council (NPC) and DOE's 1984 survey of U. S. enhanced oil recovery potential. The five models are: Polymer Flood Predictive Model (PFPM) (containing a waterflood option), Carbon Dioxide Predictive Model (CO2PM), In-Situ Combustion Predictive Model (ICPM), Chemical Flood Predictive Model (CFPM) and Steam Flood Predictive Model (SFPM). Currently, the data files for these models require tedious data entry using a text editor to prepare them for use. There are no prompts for the user, nor data validity checks to help build the data files. This requires the user to have knowledge of each model's utilization of its data file to insure correct assignment. The student or engineer who does not deal with the model on a day to day basis may find that building the data files is so laborious that he may not proceed with the run. An interactive data entry program has been developed for an IBM PC or compatible microcomputer to easily create data files for use with these EOR predictive models. It is written in the C language and employs a commercial screen utility package. The program facilitates entry of variables ranging from reservoir characteristics to economic tax parameters. New data files can be created or existing ones edited. Context-sensitive help is available and messages appear on the screen to aid the user in the data entry process. The program allows the user select items from pop-up menus. enter default values, or input new values. - Entries are checked against ranges of values to comply with the models' requirements where applicable. The data file created with this program is ready for transfer to the mainframe for use in the ennanced oil recovery predictive models. No knowledge of the required file format is needed to create it. CURRENT SYSTEM The mainframe computer is an ideal tool for small simulation projects or predictive model runs. Uncertainties and variabilities in both reservoir and economic conditions can be analyzed quickly. P. 211^

SPE Members ABSTRACT: Optimization methods of linear and non-linear programming have been used in the oil industry to study the scheduling of oil well production and drilling, and to determine the optimal investment policies. There are few studies published in the area of optimization of well stimulation policies. This paper describes the application of non-linear mathematical programming methods to optimize the design and evaluation of oil and gas wells for stimulation. The model presented is modular and it has three sub modules: SCREEN, DESIGN, and OPTIMIZE. The screening module combines pre-stimulation production performance data, geological and reservoir data to match a well to one of four stimulation operations, namely, hydraulic fracturing, matrix acid fracturing, acid fracturing and recompletion. In the second module, the design of the stimulation operation selected from the screening module is performed. The program uses widely accepted equations and correlations to design hydraulic fracturing, matrix acidizing, acid-fracturing and recompletion. In the optimization module, analytical decline curve and economic discounting techniques are used to predict the production profile after the stimulation operation, and to determine the time value of money and production. The optimization model combines the reservoir and economic data to determine the optimal well stimulation strategy. The constraints of the optimization model, include those imposed by the reservoir, the production facilities, operating costs, and operating budget of the field. Program development and validation procedures, and the application to field data are presented in the paper. Introduction Optimizing well stimulation deals with the problem of allocating resources among competing stimulation techniques to maximize production and ultimate recovery. It requires the prediction of cost and the corresponding well performance resulting from the different well stimulation strategies. Optimization of well stimulation requires a systems approach to be meaningful. Not only are the reservoir aspects of the formation evaluated but we must also consider the interactions with other sub-systems like production facilities and economics. Optimization depends greatly on the type of stimulation treatment and format-ion parameters, and consequently, the number of significant factors could range from a limited set to one including a large number of variables. Langenkamp defines well stimulation as a technique of getting more production from a downhole formation. Three points are suggested by this definition. First, well stimulation involves a technique. This could be any of the treatment models discussed in this study. Second, a reservoir that reacts to the effect of the technique is needed. The third is the interaction between the reservoir and the technique. The relationship between the technique, reservoir, cost and constraints provide the basis for optimization. To develop an optimization module to evaluate stimulation, three basic model requirements must be satisfied. The first requirement is a reservoir model that determines the production rates and recoveries given a stimulation treatment. There are two ways of characterizing a reservoir for this purpose. The first and more sophisticated method is the use of a finite difference reservoir simulator. The simulator is run and the production rates and recoveries resulting from various stimulation schemes are evaluated. The second method is to use production decline curves to describe the reservoir. The latter has advantages in some ways. It is simpler, requires less computer time, and lends itself to the development of simple closed form analytical solutions. P. 149^

Abstract A method to numerically solve the two-parameter non-linear hyperbolic decline curve equation utilized in petroleum engineering is presented. The method only requires the initial production rate and cumulative production at two points in the decline trend to calculate the decline curve exponent 'n' and the initial decline rate Di. The proposed method eliminates the need for a trial and error estimation of 'n' or the need for type-curve or graphical analysis to determine the value of 'n'. This paper demonstrates that the decline curve exponent 'n' lies between negative infinity and positive infinity and includes the exponential (n=O) and harmonic (n=1) decline cases. The results of the application of the method to both published and field data, and a comparison of the results is presented. As an adjunct to this work and for comparison, a two-dimensional, non-linear least squares regression analysis was used to calculate the two parameters, 'n' and Di. The regression analysis was based on a GAUSS-NEWTON type iterative technique which is capable of generating its own initial values. The results of using this regression analysis approach to the determination of 'n' and Di are in excellent agreement with the new numerical solution. Introduction Petroleum engineers are frequently faced with the problems of determining remaining reserves, and the prediction of future production rate. Decline curve analysis has been the tool most commonly used for this exercise. There is a considerable body of literature on the subject of production decline curve analysis. Almost all of these use the rate-time and rate-cumulative, two-parameter relationships proposed by Arps. Arps's decline curve equations of loss ratio is also the basis of the work discussed here. P. 207^

SPE Member Abstract A microcomputer program has been developed for determining the pressure drop caused by the pneumatic transport of drilled cuttings in air drilling operations. The program helps in estimating the volume requirement of air for such operations. The results of the microcomputer program are compared with other calculation methods, experimental data and field data. The comparison establishes the validity and usefulness of this model. Introduction Air (gas) drilling methods are mainly used in drilling offset (production) wells in formations where geology is known and pressure is low. It is estimated that about 10% of domestic as well as international on-shore drilling of oil and gas wells use air and gas drilling techniques for the major length of the hole. Studies have reported that there is strong evidence that air/gas drilling could be utilized to drill about 30% of domestic land wells. Air drilling has also been used in some offshore drilling (such as when drilling through reef formation and loss-of-circulation formations). Future new application of this technology is in drilling wells for enhanced oil recovery operations which generally requires multiple holes (with a minimum of formation damage). The advantages of using air as a circulating fluid are substantial. Rates of penetration are significantly increased (three to four times) as compared to conventional mud drilling methods. Fewer bits are used when drilling with air/gas. Many wellbore problems such as lost circulation and sensitive shale sloughing are virtually nonexistent. Air drilling technique makes it possible to detect the presence of hydrocarbons almost immediately. The use of compressed air/gas in a rotary drilling system has the same primary functions as the use of mud in mud drilling systems. One of these primary functions is to carry the cuttings to the surface. To ensure such functions, it becomes the main concern to provide the proper volumetric flow rate of air/gas in these drilling operations (as the proper mud flow rate determination is the major concern in mud drilling operations). As mentioned above, the most important factor to consider in setting up an air drilling operation is the volume of air necessary to do the work of hole clearing and transporting the drilled cuttings to the surface. Empirical correlations have been developed and proposed in the literature on the basis of laboratory size models. These attempts are valid over the ranges of experimental conditions investigated and over a very limited range of actual drilling conditions at best. This paper presents a microcomputer model based on basic theory of gas-particle flows to calculate the annular pressure drop requirements of air drilling operations. The minimum air volume requirement calculations can be made on this basis. The prediction of air volume requirement depends, primarily, on the pressure drop characteristics of the flow of the mixture of air and drilled cuttings (solid particles) in the annular flow stream of the borehole. The drilled cuttings transported to the surface from the bottom of the hole are of various sizes, shapes, densities and compositions; and their presence in the air flow affects the pressure drop differently. The microcomputer model presented in this paper provides the necessary tool to the designers and operating engineers for calculating optimum air injection rates in the field. The model presented here also helps eliminate the need for dependence on the purely empirical correlations for such calculations. P. 181^

SPE Members Abstract This paper presents a system modeling program, designed for use on a microcomputer, which calculates steady-state pressures and flowrates for a single-phase gas or water system. The program can model flow through 1) well completions using IPR formulas; 2) flow lines (inclined, vertical, or horizontal); 3) fixed orifice chokes; 4) and single- or multi-stage reciprocating compressors. The program uses nodes to identify pressures and node connecting elements to identify flowrates and determine which type of element that is being modeled. The program determines unknown pressures by either sequential or simultaneous methods. In the sequential method the program analyzes each node connecting element individually. The simultaneous solution creates a matrix containing flow balance equations from each node in the system. Next, the program solves this matrix simultaneously to determine values for the unknown pressures and flowrates. The purpose of this paper is to explain the logic of the modeling program and demonstrate its applicability in a microcomputer environment using examples. Introduction Surface facility modeling is an important tool in the design of new facilities as well as the monitoring of existing facilities. A good model should be able to solve the unknown variables in the network in a relatively short period of time. Second, the model should show the effects of changes in design parameters over the entire network. Several factors important in designing a good modeling program are: modeling approach, solution method, and flow equations. Steadystate flow through the network will be assumed in the construction of the model. Various computer models using the steady-state approach to facilities modeling have been available for quite some time. The program's solution method should follow the guidelines of a good model. A simultaneous solution method can solve for the unknowns in a network quickly and can show the effect on the network of changes in design parameters. Another method which can be used to solve for unknown pressures and flowrates in the network is the sequential method. However, this method will not show the effects on the network of changes in design parameters. Finally, the program should offer a wide variety of flow equations for determining the effects of different types of flow elements on the network. This paper presents a steady-state modeling program with two solution techniques available for microcomputer application. The program uses either a sequential or a simultaneous iterative technique to solve for unknown pressures and flowrates in an existing system. The sequential technique allows the engineer to solve a system much faster than the simultaneous iterative technique. This reduction of CPU time is of great importance when trying to solve large systems on a microcomputer. Another important part of the network program is the type of flow elements that it can model. P. 217^

Abstract This paper presents a model for the economic assessment of the balanced secondary recovery of infill drilling. The optimum production rate will be reached when further increase in production rate by increasing the number of wells will no longer contribute to the measures of profitability values. The assessment relies on a developed recovery-process model for reasonable and accurate projections of the production profiles of different well densities that will be the basis of financial evaluations. This model, which was designed for microcomputer solution, may be separated into two distinct parts. The first part permits the analytical prediction of the production rate profile based on recovery efficiency, which is dependent on drilling density, equipment capacity, and reservoir capacity. The second part of the model is an economic analysis of the production profile. Revenues from production and oil price profiles are combined with costs to yield a net cash flow. The net cash flow is then discounted and it results in three measures of profitability values. A sample example has been worked out. The program Listing in FORTRAN IV language can be made available for potential users. Introduction The only methods effective for large scale improvement of oil recovery are the balanced secondary recovery method of infill drilling and the application of heat in viscous oil reservoirs. Reservoirs are seldom homogeneous. Characteristics such as grain size, grain composition, sorting and diagenetic and authigenic history all play important roles in porosity, permeability, and effective drainage area. Lateral continuity is a major controlling factor in drainage area. Forrest F. Craig, Jr., gives details on how to express stratification quantitatively, but does not deal with the problem of changes in permeability caused by changes in the environment. Variation in rock quality, and uncertain reef reservoir continuity in Michigan required infill drilling to allow various porosity zones to be fully exploited within the flood pattern and to maintain a minimum well spacing for oil bank formation. Well spacing is the key to solving recovery problems caused by heterogeneity. Infill drilling will permit production of oil from parts of the reservoir that might be bypassed by standard low density well spacing. On properties that have wells of varying ages, the production characteristics of older wells should be compared with those of new wells. It is not uncommon to find wells spaced in such a way that they are in very poor communication. This results from the fact that real reservoirs are heterogeneous. Formulas and methods developed for homogeneous reservoirs, such as material balance calculations have been erroneously applied to heterogeneous reservoirs, resulting in unreliable estimates of oilin-place. Also, model study predictions are only as close to reality as the premises and assumptions used in developing the program for the model. The collapse in oil prices during the past several years has drastically altered the economics associated with oil and gas projects. During this time of lower oil prices, only companies making well thought, sound investments will survive until the next boom. A recovery process model for reasonable and accurate projection of production profile, that will be the basis of financial evaluation is needed. Well bore flow efficiency under the expected modes of production operations, coupled with a detailed reservoir description, are necessary for an accurate evaluation of production rate forecast. P. 13^