No abstract

No abstract

Asheim, Harald, SPE, Norwegian Inst. of Technology Summary In two-phase flow, both holdup and pressure loss are related to interfacial slippage. A computation model based on phase slippage has been developed that allows a priori estimation of the slip parameter values. By parameter optimization, the accuracy of the model can be improved. The model was tested with production-well data from the Forties and the Ekofisk fields and flowline data from Prudhoe Bay. It was considerably more accurate than the standard models that were used for comparison. Introduction The accuracy of predictions of pressure drop for steady flow of oil and gas in pipes is not as good as one might wish. Therefore, a new model has been formulated and implemented in a computer program called MONA. The model has two unique features. The first involves a parametric description of holdup and wall friction. The parametric description of holdup and wall friction. The holdup and wall friction are described by three independent parameters. These are related to hydrodynamic parameters. These are related to hydrodynamic phenomena and can be estimated a priori for a given flow phenomena and can be estimated a priori for a given flow situation. MONA', s second unique feature involves optimal-flow-data matching. Where flow data exist, the program can be run in data-matching mode. The program then finds the values of the three flow parameters that minimize computation errors. This enables continuous updating of the model to increase the accuracy of flow prediction and deisign computation further. Model Description Momentum Balance. The usual way to compute two-phase steady-state pressure losses is to start with the momentum balance equation for average two-phase properties. The following momentum balance is used: properties. The following momentum balance is used: (1) The two-phase densities and no-slip holdup are defined as follows: ....................... (2) ..................... (3) and .................. (4) Most models estimate the two-phase friction factor and the holdup by empirical correlations. This model determines holdup and friction factor as described below. Holdup Determination. A linearized functional relationship is assumed between gas and liquid velocities. The constants a1 and a2 can be interpreted as slip parameters: ............................. (5) The liquid holdup can be computed by a combination of Eq. 5 and a volume balance, as shown in Appendix A. This computation gives the holdup expressed by the flow parameters and the superficial velocities: parameters and the superficial velocities: .......................... (6) The assumption of a functional relationship between phase velocities is indeed equivalent to a holdup phase velocities is indeed equivalent to a holdup correlation. The advantage of phase velocities is that they relate directly to the dynamic behavior of two-phase flow, whereas holdup also depends on volume fluxes. Semitheoretical predictions of phase velocities have been made by several authors. This enables a priori estimation of the slip parameters (a1 and a2)- Two-Phase Friction Factor. The correlation for the two-phase friction factor has been derived by an extension of the two-phase similarity analyzed by Dukler et al. The extension is described in Appendix B, which gives the following two-phase friction-factor correlation: .................................. (7) SPEPE p. 221

Cresting towards horizontal wells with bottom water drive and edge water drive has been experimentally investigated using a scaled laboratory model. For bottom water drive the experimental results are compared with two models for transient crestal behavior. The range of applicability for critical rate estimates by the analytic model is found. For edge water drive the model developed predicts the measured steady-state supercritical production performance over a wide range of flow rates. As a consequence of the model and experimental results follows that by increasing the total production, net oil rates much above the critical may be obtained. Introduction Producing from an oil zone underlain by water or overlaid by a gas cap, the water-oil/gas-oil interface will deform and move toward the well. This is caused by the pressure drawdown created by dynamic flow of oil toward the well. The phenomena is called coning for vertical wells and cresting for horizontal wells. In the following we consider water cresting, realizing that water and gas cresting are governed by the same physical mechanisms. Many authors have addressed the problem of cresting, seeing that knowledge of crestal behavior is needed to optimize oil production. In the literature, the crestal performance is usually sought predicted by analytic, steady-state models. In reality, a considerable amount of the oil will be produced under transient conditions. Steady-state models may still be useful when considering oil production subject to cresting. However, their limitations should be recognized. Below, cresting behavior is investigated experimentally using scaled laboratory models. The cases considered are transient cresting with bottom water drive and steady-state supercritical cresting with edge water drive. The measured results with bottom water drive are compared to the analytic model by Giger and to a simple transient model. Supercritical cresting performance has only recently been considered in the literature, although actual wells are usually produced at rates above the critical. Thiefenthal considers the after-breakthrough production of oil and gas from Troll production tests, achieving correspondence between observed rate variations and predictions by a numerical model. The current work on edge water drive supplements the above mentioned paper by developing an analytical model for steady-state, supercritical inflow performance. The model predictions correspond closely to measurement by the scaled laboratory model.

No abstract