Facies Constraints Prediction Method For Deep Volcanic Reservoir In Songnan Area

There are multiple stages of volcanic rocks during the deep fault depression period in the southern part of the Songliao Basin.Early oil and gas exploration showed enormous potential,making it an important position for increasing oil and gas reserves and production.However,the complex genesis of volcanic rocks,rapid reservoir changes,and difficulty in geophysical prediction constrain the process of deep volcanic rock oil and gas exploration and development.Unlike the research process of conventional volcanic rock seismic exploration,which focuses on "envelope identification,insider stage,and reservoir sweetness",this article starts with the smallest research unit and first builds a multi mineral component model of volcanic rock.Combined with volcanic rock core experiments and special logging,the original state modulus characteristics of the main components are determined,and the micro connection between rock skeleton and facies controlled reservoir development is comprehensively analyzed to guide the construction of rock physical models,And lay a research foundation for Bayesian classification and normality testing of the "discrete continuous" statistical law of volcanic rock reservoirs.Then,based on the probabilistic prediction technology driven by lithofacies,research was conducted on the construction of objective functions and posterior distributions specific to volcanic lithofacies.A "facies-parameters" joint prediction scheme was proposed that can simultaneously meet the development of at least three types of lithology(facies)in volcanic rock formations,and good prediction results were achieved in the actual data of the deep Huoshiling Formation in the Songnan Basin.This article focuses on multi-scale data such as core,logging,and seismic data of volcanic rock formations in the research area,and systematically conducts research on rock physical characteristics analysis and modeling,volcanic rock physical simulation and identification,and probabilistic inversion methods for volcanic rock earthquakes(1)A Study on the Physical Characteristics of Volcanic RocksBased on the actual geological and logging data of the target layer in the research area,rock physics experiments,microscopic observations,and logging curves are used to analyze the skeletal and pore characteristics of different volcanic rocks in the actual formation.Based on regional geological background and volcanic rock diagenetic geological analysis,analyze the development and physical characteristics of reservoirs under the classification of volcanic institutions,lithofacies,and lithology,and clarify reservoir classification through well logging evaluation to guide the research of facies controlled reservoir prediction methods.At the same time,in order to better study the factors that control the storage of volcanic rock facies,rock physics experiments also include the analysis of the impact of lithology and porosity on elastic parameters,and clarify the characterization method of effective reservoir elastic parameters.(2)Physical simulation and response characteristics analysis of multi rock facies volcanic mechanismsBased on the actual geological development,a complex volcanic rock facies mechanism model was designed and physical simulation experiments were conducted.The model materials were prepared according to the principle of similarity for typical rock facies and lithology of multiple volcanic periods,and the production of fine models of volcanic structures was completed using the pouring 3D printing process.Design the acquisition system and imaging scheme according to the actual threedimensional observation system to obtain seismic data of deep and complex volcanic structures.This article conducts research on the seismic response characteristics of volcanic channels,volcanic institutions,and near,medium,and far sub equivalence of explosive phases designed around the model.It is proposed that the accuracy of volcanic rock identification can be improved by utilizing seismic texture attributes,structural tensor attributes,and other attributes.(3)Research on Physical Modeling Methods for Volcanic RocksThe lithology of volcanic rocks is complex and the pore structure is variable,requiring better restoration of mineral component content and modulus parameters to achieve the construction of multi mineral average skeleton model and saturated fluid model.The mineralogy of volcanic rocks confirms that there are significant differences between diagenesis and sedimentary rocks during the process of magma eruption and cooling,resulting in the complexity of mineral geochemical composition and geophysical laws.In this study,special logging methods such as element capture logging and nuclear magnetic resonance logging were used to study the skeletal components and elastic modulus of volcanic rocks with different lithology through standard wells.The elastic modulus changes caused by the structural differences of minerals in volcanic diagenesis in sedimentary rocks were analyzed and calculated to determine the elastic modulus database of volcanic rock minerals in the in-situ state of the study area.(4)Research on Probabilistic Inversion Method for Volcanic Rock EarthquakeThe physical model of volcanic rocks can solve the characterization of complex lithological patterns at the micro scale,while the diversity of reservoir control factors caused by the complex volcanic genesis also restricts earthquake prediction research at the macro scale.The complex seismic reflection information inside volcanic institutions requires the establishment of geophysical models with multiple lithofacies and lithology of volcanic rocks to improve the accuracy of reservoir prediction.Statistical analysis shows that the physical parameters of volcanic rock reservoirs exhibit a probability distribution characteristic of "multi kurtosis".Therefore,a "multi kurtosis" discrete probability model is studied,considering that the prior probability of the model to be inverted follows a mixed probability density distribution,so that the prior probability "kurtosis" is the same as the classification of reservoir lithology,which can not only meet the requirements of volcanic rock facies controlled reservoir prediction,but also better extract effective information from seismic reflection information,thereby improving the accuracy of complex volcanic rock reservoir prediction.The following innovative work has mainly emerged in the research:(1)A rock physics model suitable for volcanic rock reservoirs in the study area was constructed,and based on quantitative prediction of mineral components,the rock physics model was applied to determine the elastic parameters of the main controlling minerals in the reservoir in situ,providing a basis for analyzing the special geological origin of volcanic rock minerals,and providing accurate mineral parameter information for accurately predicting the elastic parameters of volcanic rock reservoirs in the study area;(2)We have established a more comprehensive seismic probabilistic inversion framework for volcanic rocks that combines lithofacies and parameters,including key links such as the division of discrete lithofacies of volcanic rocks and the normality testing of parameters.The former is an important basis for subsequent work such as Bayesian statistical inference and establishing objective functions;The latter is the theoretical guarantee for the explicit formulation of the posterior distribution formula.(3)A reservoir prediction technique constrained by volcanic rock facies has been proposed,which is suitable for complex formations with coexisting explosive,overflow,and volcanic sedimentary rock facies combinations in volcanic institutions.It achieves synchronous prediction of volcanic rock facies boundaries and physical parameters,and has achieved good application results in predicting the effective reservoir distribution of dense and basic volcanic rocks in the deep layers of the Songnan Fault Depression.