Experimental Simulation Of CH₄ Hydrate Formation And Dissociation In Silty And Sand

Natural gas hydrate has been considered as one of the alternative energy sources of the future due to their huge reserves in the ocean and permafrost zones,it is of great significance to learn and grasp the law of its formation and dissociation.In this paper,the formation and dissociation characteristics of the methane hydrate in the sediment were investigated by a self-designed high-pressure visualization experimental setup.Besides,the influence of the types of the sediment,the particle sizes of the sediment,the initial water saturation,the cycle numbers and the accumulation modes were examined.Based on the experimental results,the RBF neural network prediction model of the formation and dissociation of the natural gas hydrate was established by using MATLAB software.Firstly,the experiments were conducted in the natural sands with a particle size of6.5-355μm（sieving into four groups of different particle sizes）.The experimental results show that reducing the sand grain size（6.5～48μm）,increasing the formation pressure,decreasing the initial water saturation and repeating the formation are favorable to the saturation growth of the hydrate.When the saturation of the hydrate ranges from 10.5%to65.7%,the maximum formation rate appears in experiment 1.4 mol·s^-1·L_r^-1（L_r:unit volume reservoir）,with the particle size of 6.5～48μm.The hydrate in the natural sand exists mainly in the form of particle inclusion,pore-filling and nodular.The formation process is accompanied by the dissociation of hydrate due to the obstruction of heat transfer.The gas production curve shows an S-distribution by thermal stimulation dissociation.When the nodule-type hydrate appears in the sand,the gas production curve shows a step shape,and the maximum gas production rate is 15.2·L_STP·s^-1·L_r^-1(L_STP:litre in standard state).Secondly,the multiple formation and dissociation experiments of the methane hydrate were conducted in the silty-clay from the South China Sea.The results show that both the lower initial water saturation and the repetitive formations are favorable to the growth of hydrate.When the hydrate saturation varies from11.2%to 60.9%,the maximum formation rate is 0.2 mol·s^-1·L_r^-1.The hydrate in the silty-clay is mainly in a form of disseminated and thinly covered type.The maximum gas production rate is 6.7 L_STP·s^-1·L_r^-1 when the hydrated dissociation is accelerated by thermal stimulation.The mass transfer resistance of hydrate is estimated to be reduced due to the fissures in the sediment occurred during the gas-water transport in the process of hydrate dissociation.It was found that multi-stage the depressurization was more beneficial to prevent the hydrate reformation.The maximum gas production rate is 40.2 L_STP·s^-1·L_r^-1.Due to the high gas production rate,the intense gas-water transport causes the collapse-formation of the channel fracture to alternatively occur in the sediment.Subsequently,the formation and dissociation of CH₄ hydrate in the sediments with different stacking methods（positive stacking,negative stacking and mixed media）were investigated.It was found that the hydrates were firstly formed in the large-grained sands,mainly in a pore-filled form,while forming nodules in the small-grained sands,regardless of the stacking methods.In addition,the hydrates were also firstly dissociated in the large-grained sand layers during the dissociation process.In the experiments of mixed media,it was found that the higher the sand content,the greater the saturation of hydrate forms,and the hydrate in the mixed media is mainly of layered covering type.A step-wise gas production pattern was observed during the thermal stimulation dissociation（the initial hydrate saturation is 48.7%）and the depressurization dissociation（the initial hydrate saturation is 63.6%%）.Compared with thermal stimulation dissociation,the depressurized dissociation is faster in gas production and greater in the rate of gas production as well,with the maximum gas production rate of 51.2 L_STP·s^-1·L_r^-1 at the beginning of the dissociation.At last,the RBF neural network prediction model of the natural gas hydrate formation and dissociation was developed by using MATLAB software through the known experimental conditions and the obtained experimental results during the natural sand.The test results show that the prediction error of the model is less than 1.32%in the saturation of the hydrate and the amount of formation.The predicting deviation of the average rate of the formation is less than 9%.The relative predicting deviations of the hydrate dissociation predictions,the time of dissociation and the average rate of the dissociation are lower than 0.8%,while the deviation of the maximum dissociation rate is 12.40%.