Modelling method allowing prediction as a function of time of the detailed composition of fluids from an underground deposit in production

摘要

The deposit is divided into discrete units by a net, with each link including one or more phases of which at least one is non aqueous. A Black Oil (BO) type model is used to determine the variation of thermodynamic parameters of the phases, with supplementary input data from the thermodynamic zone of an explosion. Modelling method allowing prediction as a function of time of the detailed composition of fluids produced from an underground deposit during production in at least one thermodynamic zone defined in the model. The fluids are defined by a detailed representation with Nrd components and/or pseudocomponents. At least one function state with n parameters is determined to allow simulation at least of the preparation phase of input data, the thermodynamic behavior of the fluids over the thermodynamic route. The thermodynamic behavior of each phase and the inlet data are interpreted to produce a BO type thermodynamic representation which is completed using the supplementary input data for an explosion. A BO type model is then formed to allow the thermodynamic characteristics of each non aqueous phase to be determined in each link of the net and to give data representing the displacement of the phases in the deposit. An explosion is carried out to obtain the detailed composition of the fluids in each link. The thermodynamic route is defined by determining the variation of composition of the fluids and the properties of the different phases. The thermodynamic parameters include pressure, temperature, the saturation of the liquid and gaseous hydrocarbon phases, injection or production flow rates, and for each pair of links, the volume flows of liquid and gaseous phases. The explosion allows the determination of the equilibrium constants from input data specific to the explosion and a conversion of the results expressed in volume to give results which can be used in conservation of mass or mole equations. Each function state is used to generate n+1 supplementary functional relationships in the form of data tables or correlations included in the input data.