The Well Logging Evaluation Study Of Hongtai Gas Field

With the further exploration and development of Hongtai gas field, the difficulty of exploration anddevelopment increases. It’s urgent to solve some problems, such as the research on geologic characteristics,the determination of petro physical lower limit value for the effective reservoir, logging identification ofdifferent kinds of fluid and reserve estimation. In this paper, a system applying logging methods to identifygas reservoir and estimate reserve is formed in order to provide support for the later development of gasfield.In this paper, we firstly study the geology of Hongtai and make sure of stratigraphic and sedimentarycharacteristics of this area, then we study and analyze4-property relationship of this area. Based on theresearch of4-property relationship, several quantitative interpretation models of sandstone reservoir areestablished, such as the median grain diameter, the shale content, porosity, permeability, water saturationand bound water saturation model. These models provide theoretical basis for logging identification offluid.Based on the study on logging response characteristics of gas, the displacement pressure inflectionpoint method, porosity-permeability crossplot method and tail flick method are applied to determine petrophysical lower limit value for the effective reservoir. The lower limit value of deep lateral resistivity,acoustic travel time and the upper limit value of density are also determined. Several methods to identifydifferent kinds of fluid are applied in this paper, including cross plot chart method, three-porosityidentification method, fluid identification index method and equivalent elastic modulus difference andradio method. Based on the above research, the logging explanation standard of Hongtai area is determinedin this paper.Multi-parameter comprehensive analysis method is applied in this paper in order to forecastproductivity. We firstly fetch parameters that have good correlation with productivity, including data oflogging and gas logging curves. Then the multivariate is applied to establish a model productivity forecast.Probabilistic neural network is also applied to complete the work about classification forecast ofproductivity. Finally we determine the logging suite for the research area. The actual application shows that the methods in this paper have a high compliance rate with thelogging interpretation. So the research results in this paper have an important evaluation and significantapplication value.