The growing concern for a fail-safe method of detecting and isolating pipeline ruptures or major leaks can be attributed to a number of reasons. The justifications for considering and installing line break detection devices in the TransCanada Pipelines System can be categorized as follows:
Security of throughput for a complex multi-looped system in case of a line break
Public Relations From these, the basic operational requirements of a line break detection system are defined, namely:
Swift identification and isolation of the rupture section for further line break control measures
Pin-pointing the rupture location
Providing information as to the viability of the remaining system. It must be understood that a line rupture detection device would not improve the reliability and integrity of a pipeline transmission system.
It is an ‘after the fact’ device that is neither preventive nor remedial. However, if it is used in conjunction with other equipment and properly implemented, it can alleviate some of the adverse effects of a rupture. The purpose of this paper is to briefly outline some of the steps that TCPL has followed in the selection and, particularly, the method of testing the line break detection device presently installed in certain sections of its pipeline system.
Since rupture detectability depends very much upon transmission system configurations, operating transients and operating philosophy, it means that the selection of any detection system must be based upon a thorough analysis of the pipeline system. The key words for a detailed systematic study of transient response are ‘Sensing Parameters’. Because of the complex transient flow response in the event of a line break, particularly in a looped system similar to that of the TCPL pipeline arrangement, it is imperative that a means be available to identify the fluid parameters most sensitive to line break conditions. To this end, dynamic modelling can be used to uncover the transient behaviour and - to define the parameters most characteristic of a line rupture. From such a simulation, the selected sensing parameters can be used to locate and configure the rupture detection system. Furthermore, dynamic simulation can also provide additional information related to the effect of false alarms or false closures of mainline block valves if the detection system is connected to some triggering relays. After all, a false closure could be equally hazardous as a line break. As with any other theoretical modelling, the accuracy of this dynamic model needs to be evaluated against measured effects. These strongly suggest that ‘Rate of Pressure Drop’ or ‘Rate of Flow’ would be a powerful set of sensing parameters. However, among other problems, the main concern of using these two concepts as the basis of rupture detection systems is the vulnerability to false closure. This vulnerability usually increases with the distance between mainline valves since the rate of pressure drop increases with time and distance.