In the present article the problem of detection of leaks, which are created artificially in pipelines is examined. There were analyzed various methods of petroleum collection from a pipeline and how mechanisms of it influence the efficiency of various leak detection algorithms. Here results of modeling of various variants of collection are produced and it is shown to what changes in technological parameters of a pipeline it leads. On the basis of big experience gained by authors when applying the leak detection systems to fight the malefactors on existing pipelines, ways of development of leak diagnostic systems are determined.
Pipelines for pumping over the petroleum, gas and petrochemical products stretch on dozens thousand kilometers. Like blood vessels in an organism they run through territories of many countries. Like blood vessels they carry energy and fill economy with life. Any interruption of movement inside them is fraught with serious problems and fatally dangerous. Pipelines are considered as one of the most safe and reliable types of transporting petroleum. This is true, their work does not depend on traffic jams, on ice or suddenly dropped out snow, on storm or hurricane. Unlike tankers, they do not threaten to cover coast with petroleum, and do not create threat of a fire on a highway, as gasoline tank trucks. Reliability of pipelines is ensured by their design., The intellectual leak diagnostic systems are protecting the environment. They are based on the advanced measuring tools, which examine pipelines twenty four hours a day and will stop pumping over at any threat which in real life never arises. When we started dealing with problems of leak detection in pipelines, it seemed it was so. What will happen if a pipe bursts was possible to study on mathematical models or in the process of specially organized experiments. If within several years there was one leak, we were happy. Record of change in technological parameters during real leak was very valuable. By the moment when our systems were controlling about 8000 kilometers of pipelines, we had a big database of experiments and only three - four real failures. Unfortunately, these times have passed. For last two - three years our database was rapidly replenished with huge amount of data. We are aware of the pipes on which leaks happen each three days. Certainly, this growth is not connected with catastrophic degradation of pipe production methods or nature's anomalies. As well as in many other cases, this increase took place due to human activity. People drill pipes and steal petrochemical products. They do it so frequently, that all other reasons of occurrence of leak can be neglected. And such activity radically changes requirements to the leak detection systems.
Thefts of petrochemical products do not cause essential damage to the oil companies. In fact, the volume of pumping over in pipes is so great, that volume equal to some automobile tanks with gasoline or diesel fuel which can be stolen from a pipeline.
