Study On Dynamic Risk Assessment Methods With Uncertainties For Shale-gas Fracturing Equipment

Shale-gas exploitation uses the recently introduced techniques,such as hydraulic fracturing and cluster-well groups.This has spurred the exponential growth of gas well drilling and significantly improved the production of unconventional oil & gas.The development of industrial fracturing indicates that a global exploitation of the shale-gas reservoir is becoming increasingly popular.The fracturing operations involve different kinds of equipment and substances.However,considerable uncertainties exist within the fracturing equipment,including multiple stages,large volumes,and high pressures.In particular,dynamic risk factors are remarkably serious over the whole fracturing period due to complex and variable operational conditions.Therefore,this dissertation proposes dynamic risk assessment methods with uncertainties for shale-gas fracturing equipment based on its failure characteristics.The proposed methods can effectively reduce,transfer,and control equipment failure risk,and thereby improve the safety,reliability,and stability of shale-gas resource exploitation.(1)The traditional method of linear weighted sum is not sufficient to capture the non-linear relationship between risk factors.Hence a non-linear risk assessment method is proposed using the real-time fracturing parameters in this dissertation.Several risk parameters which have dynamically time-accumulating characteristics are specifically considered to establish a set of systematic and thorough assessment index system.Furthermore,the topological network is applied to deal with the correlation of indexes in the same hierarchy levels.The value function is also used to express the non-linear index relationship among different hierarchy levels.Finally,the overall risk,failure buckets,and risk levels of fracturing equipment are dynamically assessed.Results show that the proposed method is more reasonable.(2)Quantitative risk assessment which is a common technique cannot sufficiently monitor the real-time failure states of fracturing equipment over a period of multi-stage sand fracturing.Hence,a matrix-based risk assessment method is presented to improve conventional technique using stress-strength interference theory and value function to calculate static and dynamic failure probabilities of fracturing equipment,respectively.Further studies are developed to integrate these two probabilities with the consequences of equipment failure,particularly for the establishment of risk assessment matrix.Lastly,the failure risk states and levels,as well as risk characteristics of fracturing equipment are clearly delineated.Results indicate that the presented method is more detailed.(3)Existing classification techniques are not adequate as they do not consider the epistemic uncertainties from the assessed objects,assessment procedures,and assessor subjectivities.Hence,a risk level assessment method of equipment failure is proposed using extension and matter-element theory.Based on the preliminary risk classification,the correlative functions are firstly established to eliminate the epistemic uncertainties.Moreover,the acceptable risk levels of fracturing equipment are assessed in a dynamic and quantitative way,and the real-time alarms are also conducted.Results indicate that this proposed method is more practical.(4)Functional failures frequently occur in the fracturing equipment.In addition,main functional states are assumed to be fuzzy.Therefore,a new functional-failure risk assessment method using functional-failure identification and propagation model and fuzzy logic system is put forward.The behavioral rules and functional failure logic are expressed as knowledge base and inference engine,respectively.Fuzzy logic system is particularly applied to analyze the fuzzification and defuzzification of functional states.Finally,the failure propagation paths are effectively provided and the functional-failure risk levels are quantitatively assessed.Results show that this method is more accurate.