A NEW METHOD TO EVALUATE RESERVOIR PORE STRUCTURECONSECUTIVELY USING NMR AND CAPILLARY PRESSURE DATA

摘要

It is very important to understand reservoir porestructure information in finding interested zones andformulating advisable recovery scheme. Capillarypressure data is critical material in evaluating reservoirpore structure, from which, pore size distribution andpore structure parameters can be acquired, and by which,the quality of reservoir pore structure can also beidentified.With the analysis of hundreds of capillary pressurecurves, it is found that they all have the same mercuryinjectionpressure, so the differences in shape and typeof capillary pressure curves are controlled by themercury-injection saturation. For capillary pressurecurves with better quality, the mercury-injectionsaturation is relative higher in the same mercuryinjectionpressure, and the reservoir porosity andpermeability are also higher, vice versa. Based on which,we develop a novel method to construct capillarypressure curves consecutively integrating reservoirporosity and permeability. Correlations among mercuryinjectionsaturation and porosity and permeability inevery mercury-injection pressure are established. Withthis correlation, the capillary pressure curves can beconstructed.Generally, capillary pressure curves are exhibited insemi-log coordinate, mercury-injection saturation islinear in X-axis and mercury-injection pressure islogarithm in Y-axis. If we display them in log-log axis,they will be hyperbolic curves. The ratio betweenmercury-injection saturation and mercury-injectionpressure in the inflexions of hyperbolic curves arenamed Swanson parameters, which have a goodrelationship with reservoir synthesized index, fromwhich, we can derive reservoir permeability.The problem of evaluating reservoir pore structureconsecutively is the limitation of quantity of capillarypressure data because of the expensive test cost. As isknown to all that NMR T_2 spectrum contains plenty ofinformation of pore structure, porosity and permeability,we present a model to extract Swanson parameter fromNMR T_2 distribution. With this parameter, we can gainreservoir permeability, associating with NMR totalporosity, the reservoir capillary pressure curves can beconstructed consecutively. Comparison with thecapillary pressure curves of mercury-injectionexperiment demonstrates that they are commensurate.At last, an in-situ example is exhibited of evaluatingreservoir pore structure using constructive capillarypressure curves consecutively by above method, whichshows that it is available and accurate in detecting thechange of reservoir pore structure as a function of depth.