A Supply Reliability Assessment Study Of Natural Gas Pipeline Based Upon Demand Uncertainty

In the reliability assessment,the variation of gas demand in an assessment period is an important and basic data for the study of gas supply reliability,which is usually deterministic forecast result.These deterministic forecast results usually do not consider the stochastic property of gas demand,which will inevitably affect the objectivity of the supply reliability assessment,nor to guide the operation of the gas pipeline.In this paper,a new gas supply reliability assessment method based on max flow algorithm is proposed to deal with stochastic gas demand.First of all,by determining the influence factors and using wavelet decomposition,a gas demand forecasting model of neural network is proposed.By analyzing the error distribution of the model,the stochastic demand was transform to a normal distribution.Secondly,analyzed the transient process of user-adjacent pipeline while demand changed,the result proved the feasibility of optimizing the gas supply capacity of network under the steady state.Thirdly,considering the gas demand stochastic property and the important degree of users,to maximum the probability of fulfilling users' gas demand as the objective function,a mixed integer programming model was proposed to optimize the gas supply capacity of network.Based on the model,a gas supply reliability assessment method with reliability and availability under users' satisfaction was proposed.Finally,a virtual natural gas network case was study by using the reliability assessment method,which validates the feasibility of this method.The result shows after considering demand stochastic property,the supply reliability decreased,and with 10% forecasting error decrease,the supply reliability is increased by 6.82%.