Gas Condensate Well Test Analysis Of Variable Rate

Condensate gas well test method of variable flow can obtain more accurate stratigraphic information; however, PDG data is large and contains lots of noise; condensate gas reservoir shows retrograde condensation phenomenon; PDG data is difficult to fit and need more data. The paper carried out denoising processing on actual PDG data using db4, sym4, coi2, bior1.5 wavelet functions, certained db4 wavelet with unbiased likelihood threshold is the best plan though comparison of each wavelet functions and threshold.methods. The paper found that decomposition levels of wavelet function is the main factor affects the compress level of PDG data though analyzing of different wavelet functions and different decomposition levels on compress PDG history data. Although deconvolution method can obtain more information from PDG data, the existed deconvolution algorithms aren't adapted to pressure derivative curves with complex shape. To cure this problem, the paper proposed deconvolution algorithm based on NUAT B-spline, and experiments demonstrate the advantages of NUAT B-spline deconvolution algorithm adapted to complex well test curves; around NUAT B-spline deconvolution algorithm, the paper studied numerical inversion methods and methods to deal with production history, analyzed and discussed advantages and applicability conditions of deconvolution algorithm though application research of different reservoir data. When the bottomhole pressure reduced to below the dew point condensate gas well will appear condensate clogging phenomenon, which makes fluid flow characteristics of condensate gas reservoir occurs complex changes. The paper constructed unstructured hybrid grid for numerical simulation, and studied the influence of condensate plug and capillary number effect on condensate saturation distribution. As a basis, the paper further built and solved single well, interference well testing and multi-layer numerical component models, studied the influence of various parameters on the bottom hole pressure dynamic and bottom hole pressure dynamic characteristics under multi-channeling case, analyzed well test interpretation methods focused on single layer and multi-layer condensate gas models. The paper analyzed advantages and disadvantages of well test automatic matching features, genetic algorithms and simulated annealing algorithm. On this basis the paper proposed genetic simulated annealing algorithm for well test automatic matching. This algorithm makes up for the disadvantages of using genetic algorithms and simulated annealing algorithm been used separately. And many examples proved that genetic simulated annealing algorithm has greater global and local search abilities, and can obtain better average value and standard deviation, and has great advantage at the time consuming to obtain the optimal value. The study of condensate gas well test of variable flow improved pre-processing effect of PDG data, enhanced the application ability of deconvolution algorithm, completed analysis methods of condensate gas well test. This research has significant scientific and academic values and has a broad application prospect.