10 different reasons to get your boundary conditions correct and inclusive of a complete piping system when doing pipe stress analysis.

1. B31.1 Power Piping Code says so on Page 32, right before it goes into Stress Due to Sustained Loads.

• B31.3 defines a piping system as interconnected piping

• B31.3 states it more clearly here…

• I might say something about the stiffness matrix of the piping that you want to analyze should be calculated considering anything that affects the stiffness matrix (like a boundary condition that has a stiffness, restraint and/or movement).
• Or that considering thermal expansion/contraction, the pipe is happy as can be when set free and will show no displacement stress range stresses, but when you stop that pipe with a restraint that restricts its movement you will start to see stresses (especially when stopped/stiffened in the axial).
• My boss (a pipe stress engineer for more than 45 years) would say think of several connected rubber bands and their flexibility and stiffness when used as an inherent system (with each end held in your hands). Now if you make them one rubber band and you cut out one part of it to analyze, you haven’t included the effects of the other parts. They cannot be ignored. The part you cut out is not held in place by the parts that were not cut out. You cannot accurately analyze the piece of the rubber band without including the effects of the other parts.
• Often a retrofit is connecting to an existing pipe and therefore there is stiffness, movement, and transfer of loads that all have to be considered.
• Also, boundary conditions are often equipment. With equipment one has to consider the thermal growth of the equipment, the loads on the nozzles of the equipment (in order to meet allowables), and consider the equipment as rigid (unless one has compelling data showing what the stiffness of the equipment or anchor actually is, one cannot create fiction).
• Remember that slapping an anchor on the end of a tie-in is not always the worst-case scenario, and it is most probably fiction. An anchor does not account for movement or the interaction between pipes and what is connected to them. This type of consideration has caused issues in the past.
• As pipe stress analysts, we consider the worst-case conditions in the most conservative manner. We should only retreat from that position when given compelling data and/or code options to do so.

However, there are times when our back is against a wall and we just can’t get the data and/or permission to analyze (for instance a retro-fit and its existing piping). One can build scenarios or go to a tested and analyzed distance away, run, analyze and document these, but this is not ideal.  Remember that scope is one thing, and a proper analysis is another. Use the above 10 reasons to try to get what you need in order to have a thorough analysis that includes all the correct boundary conditions.