Research Of Hydrocarbon Pollution Detection Method Based On Low-Field NMR

As the incremental demand of hydrocarbon material mainly petrol and diesel oil in the development of the national economy and life, the hydrocarbon pollution accidents, especially the water pollution becomes worse. It is one of the key research object for the geophysics worker and environment protectors to detect hydrocarbon contamination simply, efficiently and rapidly. At present, hydrocarbon contamination detection includes gravimetric method, turbidity method, chromatography, resistance method, pyrolysis method, ultraviolet absorption spectrophotometry, fluorescence spectrophotometry, photoacoustic spectroscopy and infrared absorption spectrophotometry, etc. Some of these methods need standard oil, some need to liquid extraction to detect the samples, they are sampling measurement. Surface nuclear magnetic resonance（NMR） technology has becomes a routine method in groundwater exploration for its characteristic of non-invasion, high efficiency, and unique explanation. NMR has also been widely used in well logging. In this paper, based on this innovative idea, magnetic resonance detection method to detect hydrocarbon contamination is explored using a low field magnetic resonance instrument for petrol and diesel oil mixed with water in the laboratory.In this paper, the experiment is carried out under the 0.5 T magnetic field to solve the problem of low signal intensity, poor anti-interference ability and low resolution in the geomagnetic field. To detect hydrocarbon accurately, the simultaneous iterative reconstruction technique is selected for the experimental data explanation combined with the characteristics of the signal. The composition of echo signal is introduced in detail and also is expressed by matrix. In the beginning of the inversion, an initial value is given by rough calculation to reduce the number of iterations and improve inversion accuracy. The inversion program is written by matlab language. Subsequently, we construct the T₂ relaxation spectrum, and the echo signal is introduced by forward. The new inversion result using different inversion parameters is compared with the constructed T₂ relaxation spectrum to determine the appropriate inversion parameters. To check if the algorithm can process the experimental data accurately and effectively, the echo signal is added Gauss white noise to evaluate the inversion algorithm. In the experimental part of hydrocarbon pollution detection, different hydrocarbon and water experimental samples with different ratio are precisely configured. The CPMG sequence and inversion recovery sequence are selected to excite the examples and then the transverse relaxation echo signal and longitudinal relaxation echo signal can be obtained. In the last part, the spectrums of all petrol mixture samples and diesel mixture samples are analysed. According to the spectral peak position and area, the type of material is qualitatively determined. The mixture of diesel and water can be quantified and we can get a fitting equation expression, but not the petrol mixture for its complexity. The experimental method, the experimental data, the inversion method and the experimental result analysis are provided as a reference and ideas for the application of magnetic resonance technique in the detection of hydrocarbon contamination.