There are two types of production performance in the Southern-Kanto Gas Field, natural gas deposit of dissolved-in-water type. One is called “common type” which gas water ratio is almost constant during its whole producing period. The other is called “Mobara type” which gas water ratio increases rapidly and reservoir pressure is greatly depleted during its early producing period. In this study, sensitivities of reservoir parameters such as depth, drainage area, original gas in place, et al. affecting gas production performance are analysed numerically and actual production performance of “common type” and “Mobara type”-well is simulated. These studies suggest that existence of free gas initially in the reservoir and strength of water influx into producing wells cause mainly difference of production performance between “common type” and “Mobara type”.
The successful performance of the CSS (Cyclic Steam Stimulation) process at the PCEJ Pilot is largely dependent on the fracture configurations in the oil sands, McMurray Formation. From a simulation point of view, the key is how well such fracture configurations can be described in a simulation model. This study concludes that the coupling of a reservoir simulator (THERM, version 2.4, SSI) to a fracture simulator (TARSIM, version 2.4, Simtech) can be used to successfully describe the recovery mechanism of the CSS process at the PCEJ Pilot. Based on the Fixed Fracture Concept, the output from the TARSIM model such as fracture lengths and failure zone widths, were used as inputs for the THERM model. Within the fracture grids, fractures were expressed by increasing both the horizontal and the vertical transmissibilities. Fracture initiation and propagations were expressed as porosity increases when grid pressures reached the 7, 000kPa fracturing pressure determined from field study. By using the above approach, a good overall history match for the PCEJ Pilot Project was accomplished.
Three-phase three-dimensional reservoir simulator was constructed by modifying the black oil simulator. Some concepts such as estimation system of the pressure drops in the horizontal tubing and multi-block completion system are introduced to evaluate the horizontal well performance. Local grid refinement system is also introduced to reproduce the accurate performances of the horizontal wells. We calculated the problems which were presented at the seventh SPE Comparative Solution Project, by utilizing this simulator and conducted good performances of this simulator.
In reservoir simulation, finite difference method has long been used, which requires small grid size around a well to account for rapid pressure change. With finite difference grid for Cartesian coordinate, reducing grid size around a well results in large aspect ratio at cells far from the well and in decreasing accuracy. To deal with this problem, local grid refinement method has been developed and used. Another difficulty in Cartesian coordinate is the grid orientation, which greatly affects flow of fluids because pressure gradient can be calculated only in the direction of the grid. In the case of horizontal well simulation, gridding is dominated by the direction of the horizontal well, which makes existing finite difference approach nearly impossible to simulate. One solution to these problems is to use flexible grid. We have developed “VERDI/2D”, which is a prototype of two-dimensional, three-phase reservoir simulator using flexible grid. In VERDI/2D, control-volume finite element (CVFE) method is used to discretize flow equation. Voronoi grid method is used to generate grid by overlaying sub-grids (cylindrical or Cartesian) on background Cartesian grid at any location and any angle. The validity of VERDI/2D was proved by comparing calculation results with those of widely used commercial simulator “ECLIPSE” of Intera Information Technologies Ltd.
The boundary element method (BEM) owes its elegance, computational efficiency and accuracy, to the existence of the free-space Green's function for the governing equation. If there is no such solution in a closed form, the BEM cannot be applied, and unfortunately, this is the case for flow problems in heterogeneous media. To overcome this difficulty, the governing equation is decomposed into various order perturbation equations, for which the free-space Green's function can be found. At each level of perturbation, the solution is obtained by the BEM and the summation of various order perturbation solutions gives the complete solution for the original governing equation. To improve the perturbation series, Padé approximants are employed, which accelerate the rate of convergence of a slowly convergent series and convert a divergent series into a convergent series. Two kinds of perturbation boundary element models are introduced: one for transient flow problems, associated with the modified Helmholtz operater in Laplace space and the other for steady-state flow problems, associated with the Laplace operator. The perturbation BEM shows its utility in streamline tracking and well testing problems in heterogeneous media. The sound mathematical foundations ensure the high order of computational accuracy.
Good quality of history matching is necessary to make numerical reservoir model reliable in reconstructing and predicting field performances. Automatic history matching (ARM) provides us with a reasonable, objective and quick way to get an optimum match. The ARM technique numerically identifies the permeability distribution of the reservoir model by minimizing the difference of the observed and calculated pressures along history. The authors developed a numerical code by combining quasi-Newton nonlinear minimization algorithm with optimal control process and practical well treatment. It was applied to numerical test problems concerning the Hydropulse testing. In each case, permeability distribution of inner-well area was successfully reconstructed by the code through matching pressure performances at wells.
Once corrosion occurs in oilfields, they would cause severe operating problems. Although deep knowledge of corrosion must be needed to solve the corrosion problems, the knowledge has not been sufficiently accumulated in Japanese petroleum development companies operating relatively small number of oilfields. In the above situation, a technical committee for the corrosion problem was organized in 1991 under the Japanese Association for Petroleum Technology to collect the information of the oilfield corrosion. By sending a questionnaire on corrosin experiences, 94 examples have been obtained from the companies. This paper presents some of the corrosion instances and general observations drawing from all the examples.
The depth of the wells at the Minami-Nagaoka Gas Field located in Niigata Prefecture in Japan is over 14, 000ft with up to 355°F of bottom-hole temperatures. The produced gas contains approximately 6% CO2, 2-5ppm H2S, and up to 200ppm chloride ion. During the workover operation on some of the wells, 13Cr tubings were retrieved and investigated from the corrosion stand point. The following conclusions are drawn from the investigation: 1. With temperatures above 248°F and the existence of free water, pitting corrosion was observed on the 13Cr tubings of the well producing 6, 700MMSCF with 866days. The depthrange of the corrosion was 300-5, 200ft. 2. Severer pitting corrosion was caused by leakage of the annular brine than that without the leakage. The depth-range of the corrosin was expanded to 330-8, 200ft. 3. Crevice corrosion was observed at the faces of thread seals. The cause would be the production fluid that penetrates through opening spaces between tubular thread seals near bottom-hole. 4. Scratching surface protective films on the tubings by, for example, wirelines caused localized corrosion.