Two-phase Flow Measurement Data Processing Based On Time Series Analysis

Wet gas is gas well product, which liquid-phase volume is less than 10%. According to flow rule, wet gas belongs to two-phase flow. On the basis of self-developed meter and experiment data, this thesis analyzes the measurement data based on time series analysis. According to the data of delamination waved flow and the data of annular flow, ARMA modules are built respectively. The comparison between the orders of these two kinds of modules shows that autoregression order p of these two kinds of models are all larger than the moving average order q, and the autoregression order of the annular flow is larger than it of the waved flow. We can recognize the flow type of the two phase flow qualitatively. At last, the data of gas-phase flow is processed using statistical analysis system(SAS), and ARMA module is built using the processed data as sample. The emphases is establishment of module order and estimation of parameters. To resolve this problem, statistical analysis system is used. After the module is built, this thesis use it to forecast gas-phase flowrate of two-phase flow. The results show that the forecast error is within the permitted error of two-phase flow measurement.