Research On The Evaluation Methods Of Pore Structure In Low-permeability Reservoir

Evaluation of pore structure is always one of the most difficult challenges of loggingexplanation in low-permeality reservoir. The macroscopic petrophysical properties vary withits complicated spacial distribution and allocation relation in microcosmic levels, as well aselectrical characteristics change by varied influencing factors.Conventional electricconduction model is not applicable in low-permeality reservoir which can result in a greaterror in explanation in well logging.This research takes Nanpu oilfield Ed2+Ed3as research object, which is based on knownpetrophysical experimental data, and analyses lithology, physical properties and theirinfluencing factors.Due to different shapes of intrusive mercury capillary pressure curve andNMR T₂spectrum, the pore structure of research object is divided into three kinds.We alsoanalyse their change law of physical properties, fractal features and well logging responsecharacteristics. Also we draw the conclusion that fractal dimension of fractal pore bodies hasa linear relation to Swirr, as well as to expulsion efficiency.In order to analyse the influence ofpore structure on electrical characteristics in low-permeality reservoir,this paper designs NMRT₂spectrum measurements and matching conventional core electric measurements inEd2+Ed3low-permeality reservoir.We find that there is a good negative exponentialrelationship between T₂time and the rate of resistance augment I.And the T₂timeλ/ρ,which is the time where matching pore spaces can be saturated by irreducible water and theweight coefficient is1in Swirr computation, is proportional to the exponent of T₂^I linearly.Because of bad application effect of Archie equation in low-permeality reservoir, thispaper employs two component of pores model in order to analyse the electrical characteristics,which divide pore system into two parts, macro-part and micro-part which is saturated byirreducible water and hydrocarbon can’t invade it.We also do parameter sensitivity analysis ofthis model.The result of lgging data processing can meet our needs.