This paper describes the practical aspects of implementing and operating a real-time transient model used by Peace Pipe Line Ltd. for batch tracking and leak detection on 1000 kilometres of a 2100 kilometre pipeline network carrying a wide range of hydrocarbon liquids ranging from ethane rich NGL's to crude oil. The entire implementation process from collection of raw data at the end devices to the presentation of Model computed data to the pipeline dispatcher is reviewed. Actual leak detection test results and batch tracking accuracies are also presented.
This paper is a follow up t o the paper "A Transient Model For Real Time Simulation of a Batched Pipeline Network" presented at last year's PSIG conference. In that paper the authors presented the mathematical derivation of a new linearized implicit model formulated in conserved variables. The intent of this paper is to show the manner in which that model has been implemented and made to work in an existing Supervisory and Data Acquisition (SCADA) environment. It is hoped that this information will be of benefit to pipeline dispatchers undertaking similar ventures.
Peace Pipe Line Ltd. owns and operates a pipeline system in northwestern A1 berta, Canada (Figure 2) which gathers and transports crude oils, condensates and gas plant liquids to refineris, cavern storage and other pipeline companies in the Edmonton area. The system consists of approximately 2100 kilometres of pipe ranging in diameter from 76 mm (3 inches) to 508 mm (20 inches). The products, which range from Non-Newtonian crude oils to ethane rich NGLs, are batched through the system with numerous intermediate blending points and minimum product storage.
