задача, ГРП. Оптимизация разработки участка месторождения Х путем бурения зарезок боковых стволов и уплотняющего бурения(Томская область)
 Скачать 4.23 Mb.
|
|
Geological model The geological model includes information on the geometry, structure, lithofacies composition, nature of saturation with geological and physical properties. The digital geological model is represented as the sets of blocks of properties or cubes. Researchers distinguish basic and additional cubes; where basic ones are: a cube of open porosity efficiency, initial gas-oil saturation, permeability, bound and critical water and oil saturation [1], while all other cubes are optional. To build a three-dimensional geological model, the following initial data are used: 1. Seismic survey of the territory 2. Well data (coordinates, inclinometry, stratigraphic arrangements) 3. Petrophysical research 4. GIS data on well logging The deposit framework is a collection of cells that are subsequently filled with properties. When constructing the simulation area in the horizontal plane, the size of the cells is selected depending on the size of the deposit and its type. In the vertical plane, the size of the cells is chosen to reflect the vertical heterogeneity of the formation as much as possible. As a result, the smaller the size of the cells, the better it is possible to reflect the structure of the deposit; however, the calculation time of the model itself is greatly increased, which, perhaps, results in certain simulation technical contradictions. At the next stage, a lithological model is constructed, and the filtration- capacitive properties (FCP) are distributed. The data on lithology and FCP obtained from the results of the interpretation of geophysical studies of wells, are projected onto the grid cells and propagated in the interwell space. The final stage of the geological model construction is its reliability assessment. For the most part, this stage consists in comparing the values of the 101 development object obtained as a result of simulation and actual values. The evaluation is performed on cross-rafts of open porosity efficiency, effective thicknesses, and by the ratio of balance reserves. The hydrodynamic simulation To create a hydrodynamic model, it is first necessary to perform scaling or upscaling of the geological model. The task of upscaling is to reduce the number of active cells and preserve the detailed geological integrity. In the resulting model, the nature of the distribution of the main filtration-capacitive, geometric and physical properties should be fully preserved. After upscaling is completed, the hydrodynamic model must be initialized. The initialization process consists in entering data on the initial equilibrium state into the model. This includes data on the composition of fluids, the position of the water-oil contact, the reference depth, etc. At this stage, the first hydrodynamic calculation of the model is made, the initial geological reserves of oil, water and gas are calculated. Then, the calculated values are compared with the hydrodynamic simulation results and expert estimates. The next step in creating a model is its adaptation. Adaptation of hydrodynamic simulation model is a procedure for validating a model, which consists in modeling past reservoir behavior and comparing it with actual historical data (History Matching). Comparison of the calculated and actual dynamics of development indicators is a mathematically inverse problem. The essence of adaptation is to compare the final results, with their apparent inconsistency, the input data change, and then the calculation is repeated. Adaptation is one of the most critical steps in the creation of the permanent geological and technological model. One of the most important aspects of such adaptation is the ambiguity of the results; in other words, several constructed models can give satisfactory adaptation, although none of them will reproduce the actual state of the formation to the desired extent. This is explained by the fact that 102 only a limited number of known variables per a very large number of unknowns are known about the modeled system. With manual adaptation, the whole sequence of actions to adapt the permanent geological and technological model is done by a specialist. After the first iteration, the engineer analyzes the results, corrects the parameter values, then starts a new series of calculations until the deviations between the actual and calculated values become less than the specified error. The analysis of the results for manual adaptation is basically the comparison of the dynamics graphs of the field data parameters and the model. During the adaptation, the initial data is adjusted. Most often, cubes of permeability, critical and related water saturation change since these parameters tend to have the greatest uncertainty. Thus, the adaptation includes the following stages: 1. Pressure adaptation: The main parameter is bottomhole pressure. It is one of the easiest to measure since such measurements are available for each well. Comparison of actual development data and calculated values is important for studying general trends in pressure behavior. 2. Adaptation of fluid flow rates: When adapting, not only the coincidence of oil production volumes is taken into account; it is important that the cumulative production is coordinated in phases, i.e., the water cut and the gas factor must coincide. There is no specific way for adapting the model, but it is possible to follow some simple but important advice. For example, M. Carlson suggests the following: 1. Choose the simplest model; 2. Try to change the parameters that have the greatest impact on the result; 3. Try to change the parameters with the greatest uncertainty. 103 Upon completion of the configuration phase, the model contains all the raw data, adapted and ready to continue the modeling process. The next stage implies compiling and calculating various forecasting development options. A number of requirements and recommendations for field development are sent from the customer. For this field, the key planned indicators, as well as some subtleties of the ground infrastructure or design features, are indicated. To increase the chance of success, several forecasting options are being developed, each of which has a slightly different development concept. Based on the results of this stage, the development indicators of all options are compared, and then the most optimal one is selected. This option builds the required maps, graphs of dependencies, and a presentation is prepared [2]. Modeling methods in hydrodynamic simulators Real simulation of EOR methods in hydrodynamic simulators often encounters technical problems associated with the lack of necessary simulator options, inconvenience of the interface often lacking information obtained from laboratory experiments and necessary for assignment in modeling programs. Obviously, full-scale modeling of EOR technologies such as prediction of fracture development, fluid inflow to a fracture from the formation, and movement of a multiphase flow in the fracture itself, or simulation of injection of various chemical reagents, is extremely difficult due to the need to know a large amount of data (geomechanical and/or rock adsorption properties, etc.), as well as the absence in most modern simulators of appropriate options that make it possible to carry out adequate simulation of technologies. On the other hand, in models it is sufficient to simulate the result of the action of these methods such as dynamic change in well productivity, deterioration or improvement of the filtration properties of individual interlayers (cells), etc. Currently, this method seems to be the most common, however, this approach also requires a large number of studies, which are not always cheap and do not give an unambiguous picture of the impact of geological and technical 104 actions. Thus, for fracturing, determining the crack length, its orientation, and dynamic behavior is rather a complicated task for experimental investigation, and in this case, it is necessary to analyze the uncertainty of the crack parameters in one form or another. For flow deflection technologies, it is necessary at least to study the dynamic change in the inflow / intake profile. Currently drilling of directional and horizontal lateral well bores along with hydraulic fracturing, physical and chemical impact, application of flow-deflecting technologies is considered as one of the most effective methods in the late stages of development. In fact, the approaches used in the well selection can be divided into two large groups: methods based on analysis of geological and field information, and direct geological and hydrodynamic simulation. When choosing wells for geological and technical actions, the development engineer of the territorial production enterprise uses all the accumulated experience to analyze the available geological and commercial information available: the results of geophysical and hydrodynamic well studies, reservoir studies, design features and well status, densities maps of current reserves and saturation, the results of similar actions in the same conditions, etc. Today, substantiation of geological and technical actions efficiency based on the use of intelligent systems analysis and forecasting is becoming widespread, among them, neural networks, Kohonen maps, decision trees, etc. However, in most cases, simplified engineering techniques and complex intelligent systems do not make it possible to give direct answers to standard questions such as "In what direction should we drill the trunk, and at what depth should we open the reservoir?", "As far as the interference of wells is concerned, will sidetrack drilling be technologically and economically profitable in the whole area of development?"," How to choose the optimal geometric characteristics of a sidetrack? "and others. In order to answer these questions correctly, it is advisable to use the most detailed deterministic geological actions. 105 It should be understood that the distribution of reservoir properties and reserves, even obtained with a deterministic approach to simulation, is quite probabilistic and should be evaluated from this point of view. Consequently, a large number of calculations are required for different entry points into the reservoir, with different orientations, depth and length of the wellbore. In this case, it may be necessary to change the operating modes of the surrounding wells in the area. When evaluating the effect, along with the initial oil rates, it is necessary to estimate the accumulated production both by the well and by the site. Naturally, such calculations cannot be performed effectively in a manual mode, especially taking into account the limited capabilities of hydrodynamic simulators in this regard. To solve this problem, an algorithm was developed and implemented in the form of a software product.); this algorithm makes it possible to: - load and modify the well logging source text file (including enlarging the computational grid and forming a group of wells in the impact area); - identify the variants of the sidetrack trajectory by setting the variation of the basic geometric parameters including the sidetrack displacement from the main track, the zenith and azimuth angle, deepening towards the top of the reservoir, the length of the horizontal section, perforation and technological limitations (yield, depression, bottomhole pressure); - start calculation (Windows or Linux OS) in a batch mode of a set of calculation variants with various sidetrack trajectories and processing of calculated variants; - form the tables and graphs in Excel. The following procedures for specified parameters, the launch of options for calculation, postprocessing and the formation of a report using the program is automatic, which makes it possible to calculate the required number of options 106 without direct expert’s involvement. Limitations arising in this sub-process are related only to the capabilities of computer technology. Conclusion The analysis of actual and calculated indices of horizontal lateral boreholes drilled in accordance with experts’ recommendations and undetected complications during drilling and operation (without overflows and cross flows behind the casing) shows the convergence of results. In this case, the term convergence is related to ratio actually obtained by the forecast for initial oil flow rate of about 80 % with a correlation of more than 60% while the convergence is estimated at 107% at flow rates from the fluid. The presented approach to working with the models in general and to planning and evaluation of effectiveness of sidetracks drilling, in particular, makes it possible to considerably reduce time costs and to use the possibilities and advantages of hydrodynamic simulation models thus improving the reliability of engineering decisions. Bibliography 1. West Virginia Geological and Economic Survey. 1997. Enhancement of the Appalachian Basin Devonian Shale Resource Base in the GRI Prepared for: Gas Research Institute. December 1997. 2. Dang, L., Peng, P., Zhou, G., Chen, Z., Phan, C. V., Walker, J., Zhao, J., Zhao, Z. & Liu, G. 2013. First Openhole Sidetrack in Deep Horizontal Well Saves Time and Lowers Cost: A Case Study. International Petroleum Technology Conference |