Sound Velocity Measurement Of Samples And Experiment Research With Simulation For Miniature Model Well

Acoustic logging is a kind of important approach of physical geography in prospecting for fossil fuel. Due to the transient acoustic pulse tracked continuously which is excited by sound point, reflects information inside the stratum, the existence and geometric property of fossil fuel, as well as gas, can be confirmed. The data based on acoustic logging also fill a crucial role as logging comprehensive explanation for stratum lithology and parameter of oil reservoir stratum. The full wave measuring of model well in laboratory environment is proved to be instructive for practice; when physical simulation and the real well make a certain proportion in scale, the experimental result probably coincide with real situation within some extent. So the full wave measurement of small model well exhibits significance in engineering.My thesis can be divided into two parts- experimentation and theoretical research. The experiment focus mainly on measurement of sound velocity in sticks sample, concrete rock sample, the density of which also be included, and of full wave in small model well; the theoretical counterpart is about the simulation of model well.Sample measurement: he predecessor of my lab has made some measurement about sound velocity in sample rock and other kinds of material. However some issues about the measurement accuracy and error analysis have been exposed later, so to improve the research above, my work concentrates on enhancement of accuracy and error calculation. The accuracy of the length measurement has been enhanced to 0.002cm, but there are some uncertain factors about wave shape timing, so on account of real measuring of sample by oscillograph, the threshold voltage of vibration start can be ascertained- fix on 0.7% of the first positive peak detected as threshold voltage. Thanks to the high accuracy of time measuring- 0.01μs, the velocity error in sample stick can be limited into 10m/s, which is a satisfied result.The measurement of rock density and sound velocity: For purpose of cooperation project with Daqing oil logging Inc, I have measured respectively the density of 28 concrete samples and P-wave velocity in them. I use electronic scale to get the mass of samples, most of which are about 200g, and due to high measuring accuracy of the scale (0.1mg), the final error is relatively low. To measure volume of samples, I use dry method to do it, because the samples are permeable for water. The error of density measuring mainly depends on the volume scaling, and since the scale of all samples are in concord, so the final calculation error is about±0.02g/cm3.Because of high accuracy of time measuring- 0.01us, the velocity error of rock samples I have measured can be limited into±10m/s, while the error of reference data provided by Daqing is much larger. To compare with Daqing data, I low the accuracy of time measuring- from 0.01us to 0.1us, by means of enhancing threshold voltage of vibration start, fixing on 1.3% of the first positive peak detected as threshold voltage, then ultimately control the P-wave velocity error within±20m/s. Comparing with the reference data of Daqing, I find out the values of density and velocity I have measured, are both a little greater than the reference data, which is on account of the variation of water saturation of samples. The feasibility of my measurement and data is both recognized officially by Daqing oil logging Inc. Besides, I also measure S-wave velocity in rock samples, and analyze the anisotropy of all samples. Finally, I draw the conclusion that 3 samples- sample hyp-1, sample hyper-2, sample middle-11, possess obvious anisotropy. Full wave measurement of small scale model well: my main work focuses on the way weakening and eliminating the influence of reflection wave in interface to the full wave field of small scale model well. For simulation of solid stratum, I reconstruct the grain of sample surface to lots of rhombuses- 1mm in geometric scale and very shallow in depth, which increase diffuse reflection in the interface, and use probes at main frequency 1.25MHz and 160KHz respectively to measure full wave of the model well. Compared with the previous sample, reflection wave in interface is attenuated dramatically, with V-wave, P-wave and stonely wave recognized and reflection wave overlapping with full wave field mildly. For simulation of soft stratum, after surface is slightly grained as solid model, reflection wave in interface is diffuse reflected, but overlaps with stonely wave seriously. Then I make the surface of soft model been deep and coarsely grained- with vertical directions grain, but the result is still unsatisfied- reflection wave in interface and stonely wave cannot be isolated. To analyze the data efficiently, I make two kinds of data, attained by 1.23MHz and 160KHz probe, been filtered by Window function, consequently find out high frequency component occupies a big proportion of the full wave. On the other hand to extract P-wave and S-wave velocity from full wave field accurately, I adopt STC to fulfill it.Simulation of model well: simulation is involved to validate the rationality of experimentation, as my data coincides with the simulation accurately. In solid stratum it is obvious to recognize three component- compressional wave or P-wave, shear wave or S-wave, stonely wave, while the interface of the reflected waves is also clear. The data consists of much high frequency composition, in virtue of the high frequency probe measuring, and various noises in the data bring on unclarity of the wave shape, however we can attain the ideal wave shape by numerical value simulation, with the data containing very little noise. For soft stratum, the P-wave and stonely wave magnitude of simulation are a little stronger than those of experiment data, while magnitude of reflection wave in interface is a little weaker than that of experiment data, the data accords with the simulation mainly. An explanation for above results is that the setting of program parameter leads to the phenomenon, which makes little influence to the final conclusion, so the accuracy of stratum experiment measurement can be testified. Through simulation, it can be proved that the arriving time of reflection wave in interface moves back owing to the increasing diameter of model well which is definitely a key to eliminate the disadvantageous influence of reflection wave in interface, and the conclusion above can be a instruction to make a model well with appropriate scale parameter.