Research On Well Testing Interpretation Method Of Fractured Horizontal Well In Low Permeability Reservoirs

Fractured horizontal well technology is the key technique for low permeability reservoirsexploitation. And well testing interpretation is the important means to obtain parameters oflow permeability reservoirs and evaluate the reservoirs comprehensively. The traditional welltesting interpretation method based on analytical model made hypothesis, simplification andabstraction to actual problems. But it is not accurate for dealing with such complex issues asfractured horizontal well in low permeability reservoirs. In order to accurately describe andsimulate the percolation characteristics and production performance of fractured horizontalwell, and deepen the understanding and awareness of the low permeability reservoirs, thisthesis has established the oil-water two-phase numerical well testing percolation model,which is derived from the percolation theory of low permeability reservoirs, and combinedwith fractured horizontal well technology and numerical well testing theory. The model isbased on black oil model and coupled with fluid in the hydraulic fracture, variable mass flowin horizontal well and nonlinear flow in low permeability reservoirs. For dealing with suchproblems as big grid scale difference from reservoirs, fractured fissure and horizontal well,grid generation difficulties, and usual grid generation method could hardly reflect the actualpercolation flow characteristic of the near well-bore and far well-bore zones, associated withinaccurate reservoir boundary description problems, etc., a hybrid grid generation methodcombined with unstructured grid and structured grid is set up. And based on this method, thenumerical well testing model with oil-water two-phase for fractured horizontal well in lowpermeability reservoirs is solved by using finite volume method. Because hybrid gridsgeneration has a series of problems such as complex structure, a huge number, slowcalculation and solving difficult, the way of grid generation and numerical calculation methodis improved by parallel programming. And a quick solving method for well testing model isfurther proposed, which enhances the computation speed observably. Simultaneously, thevalidity of the model is verified by using numerical simulation method. And responsecharacteristics of the model with different reservoirs, horizontal well and fracture parameters on pressure drop of horizontal well are discussed. To achieve relevant parameters of welltesting model quickly and accurately and reduce the influence of artificial factors, a groupingparticle swarm optimization algorithm is proposed, which is combined with traditionaloptimization method, particle swarm algorithm and parallel algorithm. This algorithm notonly accelerate search speed by avoiding the problem that the particle swarm algorithm iseasy to fall into local optimization, but also ensure the stability of the algorithm for combiningwith Gauss-Newton method. And the parallel algorithm is adopted at the same time, whichreduces the complexity of the algorithm and increases the calculation efficiency. According tothe algorithm principle, flow stages can be divided to some extent. In conclusion，theproposed well testing interpretation method for fractured horizontal well in low permeabilityreservoirs, which is combined with nonlinear percolation theory, hybrid grid generation, finitevolume method, particle swarm algorithm and parallel algorithm and so on, perfects usualwell testing interpretation method of low permeability reservoirs and has significant theoreticvalue and wide practical applications.