Gas Pipeline Rupture Detection and Control

Abstract

Automatic and remotely controlled main line valves are used in natural gas transmission pipelines to provide early shutoff of gas flow in the event of a pipeline rupture. Operating experience, however, shows that these valves and their associated rupture detection and valve operator systems are not always reliable in sensing a line break and in achieving valve closure. There are documented instances of pipeline ruptures going undetected, and of main line valves not closing completely after even a full line break. False valve closures have also occurred, causing pipelines to be shut down unnecessarily. Under sponsorship of the Gas Research Institute (GRI), a technology assessment program was conducted by Southwest Research Institute (SwRI) to define the present state of the art of automatic and remotely controlled main line valves, to evaluate their effectiveness in achieving isolation of a ruptured line, and to identify technology improvements that are needed to improve valve reliability. This study was based on a survey of the U.S. natural gas industry’s experience with line break control equipment, and upon computational modeling of typical pipeline systems to simulate the generation and propagation of pressure and flow transients created by a line break. Line break transients were also compared to the transient levels generated by normal pipeline operations (start-up and shutdown of compressors, branch load changes, etc.). Also during this study, a semi-empirical computer model was developed to calculate pipeline blowdown time as a function of break size, pipeline configuration, and operating conditions, even in cases where valve closure is delayed for some period after the line break occurs. This information can be of value to pipeline engineers and emergency response planners. Results of the technology assessment show that the primary source of unreliability in present day line break control systems lies in their inability to discriminate between a line break transient, and those generated by other pipeline operations. In most cases, automatic control valves (ACV’s) sense the rate of pipeline pressure drop (ROPD) to detect a line break. In many field applications, however, transient pressure signals caused by compressor operations and load changes are stronger than those produced by a line break. In order to avoid false valve closures which could otherwise result, sensitivity of the rupture detection systems is “backed off,” often to the point of inoperability of the ACV. Other fluid transient signals besides pressure drop are also generated in the pipeline during a line break, and these can sometimes be used advantageously to replace or confirm the traditional ROPD signal. In looped parallel pipelines, for example, crossover flow rate is usually a more reliable line break signal when crossovers are open. The resultant line-to-line differential pressure also provides a viable option when crossovers are either open or closed. In general, however, no one detection parameter is optimum for all applications. Of even more importance is the rupture sensor location. The concept of locating additional sensors between the main line valves (rather than just at the valves) provides the most promising approach for enhancing the reliability of present ACV’s and for providing needed line break information to remote controlled valves (RCV’s). In either case (ACV’s or RCV’s), rupture sensor location is much more important than valve location or valve spacing in ensuring reliable rupture isolation. However, valve spacing does affect the volume of gas blown down after shut-in of a ruptured section. This time can often be from 10 to 60 minutes for typical pipeline valve spacings.