Stochastic model of material balance for leak detection in oil pipelines

The paper discusses the basic algorithm of parametric systems for leakage detection - the material balance method. The main attention has been paid to the issues of its practical use: justifying decision making criteria. eliminating false triggering, evaluating the leakage detection system sensitivity. The material balance method is an algorithm of sequential oil imbalance analysis to detect imbalance deviations from zero. The mathematical formulation of the problem is to find the moment of change in probability characteristics of the random process under investigation. In this case, such a characteristic is the mathematical expectation of oil imbalance. The developed stochastic leakage detection model is based on the consideration of measured oil unbalance in a pipeline section as a stationary ergodic random process. We considered the problem of "imbalance" of such random process and suggested a way to solve the same from an engineering point of view. We developed a methodology for evaluating the leakage detection algorithm sensitivity depending on the technological site equipment with flow measuring instruments.