| LD10-1 gas field in Ledong Sag,Yinggehai Basin is located in the deep water area in the west of South China Sea.Huangliu Formation and Meishan Formation are two main target intervals buried at depths of nearly 4000 m with formation temperature of nearly 200℃,formation pressure of 100 MPa,a maximum pressure coefficient of 2.3,and the reservoir permeability is less than 5 m D.Due to the high temperature,ultra-high pressure,and high mud density wellbore environment,special logging tools are ineffective and can only obtain conventional logging data such as resistivity,neutron,and density logs.In addition,ultra-low permeability reservoirs are invaded by high specific gravity mud and the reservoir fluids are high in carbon dioxide(CO2),making it a huge challenge to evaluate complex fluid ultra-low permeability gas reservoirs using limited conventional logging data.The traditional logging evaluation method uses methods such as core fitting,Archie model,clay additional conductivity model,and Timur permeability model to calculate the porosity,water saturation,permeability,and other parameters of the reservoir.These methods are well applied in medium to high porosity and permeability,homogeneous formations,but as the pore structure of the reservoir becomes more complex and the permeability becomes lower,the logging response becomes weaker,and traditional evaluation methods have become ineffective.This paper proposes a reservoir evaluation method combining large-scale physical simulation and numerical simulation,which attempts to breaks through the bottleneck that conventional logging data cannot evaluate complex fluid extra-low-permeability reservoirs.The specific method is as follows:Firstly,the basic characteristics of the reservoir such as mineral components,fluids and physical properties were determined based on the core rock chip analysis data and logging test data of the target area.The lithology of the target area is mainly sandstone with a small amount of clay and limestone,the porosity is 11%on average,the permeability varies widely from 0.1 m D to 10 m D,and the fluid component is a mixture of methane and carbon dioxide.Seven large artificial cores of 1 m in height and 1.2 m in diameter,which are close to the petrophysical characteristics of the formation,are made according to the above conditions,basically covering different types of reservoirs in the target area,and a large FRP tank environment suitable for measurement with drilling logging instruments is established.The fluid is then injected into the large core,placed into the large FRP tank,and the drilling logging instrument is lifted to measure the response of the core in the FRP tank to investigate the key reservoir parameters affecting the logging curve and to clarify the logging response pattern in the target area.A digital petrophysical model is then established based on the reservoir parameters of the physical man-made core,and the mineralisation and saturation profiles of the high specific gravity mud intrusion are obtained using the theory of seepage mechanics.Simulations are carried out using Maxwell’s electromagnetic field vector finite element theory,nuclear physics Monte Carlo theory to obtain time-shifted logging responses such as resistivity,neutron,density and gamma logs.The large physical simulation measurements were also used to calibrate the key parameters in the numerical simulation,enabling this numerical simulation method to be generalized to the target area,where the neutron numerical simulation results are less than 2.1%in error with the MCNP software,the density is less than 0.013 g/cm3and the resistivity is less than 5%with the scale wells.The numerical simulation of the influencing factors of reservoir variables concluded that the Ledong 10-1 gas field has a large variation in pore structure and permeability with constant porosity due to the different fine mud cementation and ash cementation.Meanwhile,carbon dioxide(CO2)is critical under high temperature and ultra-high pressure formation conditions(200°C,100 MPa),where the density can reach 1 g/cm3 and the neutron excavation effect can be neglected.Finally,a set of numerical simulation techniques calibrated by large physical simulations is obtained.Combined with the actual time-lapse(with drilling,retesting,cable)resistivity and nuclear logging curves at the gas field,the input modelling parameters(e.g.mud,porosity,permeability,saturation,fluid components)are continuously adjusted to match the numerical modelling results and the actual logging curves to form a set of integrated logging inversion methods for calculating key reservoir parameters in the target area.In the application in the Ledong 10-1 area,the joint inversion method obtained porosity,permeability,saturation and carbon dioxide(CO2)content,with the relative error between porosity and core less than 8%,absolute error between permeability and core less than half an order of magnitude,absolute error between saturation and core less than 5%,and relative error between CO2content and test results less than 10%.The integrated inversion method significantly improves the accuracy of conventional logging data in evaluating complex fluid extra-low-permeability gas reservoirs,and provides reliable basic data for subsequent production and development. |