A study was undertaken to evaluate crack arrestor and mainline block valve (MLBV) spacing distances beyond the limits defined in the 49 CFR Part 192 for Class 1 locations for the design of a 42-inch (1,067-mm) OD arctic pipeline. The study assessed whether an MLBV spacing longer than that required by 49 CFR Part 192 for Class 1 locations can provide a level of safety equivalent to that afforded by the spacing recommended in the code. This was accomplished by comparing the hazards in terms of the volume of natural gas released over time, the potential for damage to surrounding structures, and the life safety risk to personnel and the public. The analysis was performed using the software tool PIPESAFE (version 2.20.0), which was developed for a group of pipeline operators by Advantica Technology (now DNV-GL).
A full transient analysis of the flow inside the pipeline and through the rupture opening was carried out with automatic shut-off valve (ASV) closures simulated as boundary condition changes at the locations of the valves triggered by the local transient pressure. Gas outflow rates were fed to a structured flame model that calculates the temperature distribution within the flame and the radiant energy emitted and uses the latter to determine the incident thermal radiation field in the area surrounding the rupture, the associated hazard areas and the accumulated thermal radiation dosage over time. These results were compiled into contour plots of thermal radiation intensity for different times; plots of the total area within specific contours of thermal radiation intensity for different times; and plots of the total area within specific contours of accumulated dosage. The dosage-area curves facilitate a direct comparison of the various MLBV and crack arrestor spacing options considered within this study by providing a simple means to establish if the change in spacing causes a substantial change to the affected areas for dosages up to the limits associated with specific levels of lethality to humans and for piloted ignition of wooden structures.
It was found that valve spacing has a strong effect on the time at which closure begins to affect the outflow rate. The decline in flow rate after valve closure had significant influence on the thermal radiation field, but these effects only occurred at a relatively late stage. Increasing fracture length led to considerable changes in the shape of the thermal radiation field, but the total footprint within which casualties might be expected in the event of an ignited rupture release and the severity of injuries within the footprint are unaffected by valve closure under the assumed conditions. Similarly, the damage potential to surrounding buildings was unaffected by valve spacing, indicating that increased valve spacing could be implemented in remote, low population density areas without affecting safety.