Theoretical Models Of Predicting Sand Erosion In Elbows For Gas And Multiphase Flows

Sand erosion,which can result in the environmental pollution and economic losses,is a severe problem in the processes of producing and transporting oil and gas.Gas and multiphase flows are common flow patterns in the oil and gas engineering,and the erosion magnitudes are more serious in conditions with higher conveying speeds.The understandings of erosion mechanisms in gas and multiphase flows still need to be improved.The elbows are connection elements applied in the oil and gas industry,and are very susceptible to erosion caused by sand particles that are entrained in pipelines.Hence,this dissertation focuses on investigating the erosion mechanisms and establishing theoretical prediction models in elbows for gas and multiphase flows.Annular flow,which is characterized with the high superficial gas velocity and low superficial liquid velocity,is the most common multiphase flow pattern adopted in oil and gas industry.Hence,this dissertation starts off with the investigations of sand erosion in elbows for annular flow.Firstly,a mechanistic model is developed to predict sand erosion on the symmetry plane in elbows for annular flow.The flow field is simplified by analyzing the characteristics of flow fields in annular flow.Mathematical models are developed to analyze the particle motion rules in gas core and liquid film,respectively.Sand particles are tracked until impacting on the pipe wall or moving outside the elbow.Then,the impingement information of particles are extracted and introduced into erosion formulas to calculate the erosion results.According to the comparative studies,the mechanistic model is demonstrated as an accurate and efficient tool of predicting sand erosion in annular flow.Moreover,the effects of some factors,such as the superficial gas velocity,superficial liquid velocity,pipe diameter,particle diameter,curvature ratio of the elbow and liquid viscosity,on the erosion predictions are investigated.Secondly,a probability model is established to predict the sand erosion profile of the elbow for annular flow.The probability model is a method of converting the occurrence probability of one particular result into the co-occurrence probabilities of several independent events.The erosion probability of any position on the elbow can be predicted using the probability model.Both the first collision and second collision are considered in this model.The impact velocities of particles that successfully cross through the liquid film are derived theoretically.The whole erosion profile on the elbow for annular flow can be obtained by combining the erosion probability and the erosion formula.The predictions of the probability model are examined by numerous experiments.Besides,comparative studies between the probability model and other existing erosion models are conducted.The erosion profiles on the elbow are analyzed based on the predictions from the probability model.The effects of superficial gas velocity,superficial liquid velocity and curvature ratio of the elbow on the distributions of erosion profiles are finally investigated.Thirdly,although the probability model is effective and accurate,the computational efficiency is relatively low,especially for situations concerning about the maximum penetration ratios and the most serious erosion position.Moreover,the erosion mechanisms in gas and annular flows are demonstrated to be similar according to experimental observations.In this case,a unified and explicit correlation is established to predict sand erosion in elbows for gas and annular flows.According to the analyses of erosion mechanisms in gas and annular flows,an erosion reduction coefficient and two erosion enhancement coefficients are established,respectively.The erosion reduction coefficient,which is derived theoretically,reveals the erosion reduction in annular flow compared with that in gas flow.The two erosion enhancement coefficients,which are established by fitting numerous simulation results,reveal the effects of the second collision on predictions of the maximum penetration ratio and the most serious erosion position,respectively.The comparative studies demonstrate that the computational efficiency of the unified explicit correlation is remarkably increased under the premise of ensuring the prediction accuracy.Fourthly,an integrated model of predicting sand erosion in elbows for multiphase flows is established.Based on the fluid characteristics of different flow patterns,the calculation methods of the impact velocity and impact angle are derived,respectively.According to the obtained impingement information of particles,the erosion models in different flow patterns can be established,respectively.After this,the erosion models can be integrated into a unified model based on the proposed erosion reduction coefficient and erosion enhancement coefficient.The detailed values of these coefficients vary among different multiphase flow patterns.The integrated model is examined by comparing with the experimental data.To further improve the accuracy,the integrated model is optimized using an optimal coefficient.The optimized erosion model is validated by the experimental data,and both the non-optimized model and the optimized model are compared with other existing erosion models to show their predictive ability.Finally,the integrated model for gas and multiphase flows is compared with other erosion models.