B31E

Standard for the Seismic Design and Retrofit of Above-Ground Piping Systems

This Standard establishes a method for the seismic design of above-ground, metallic piping systems in the scope of the ASME B31 Code for Pressure Piping (B31.1, B31.3, B31.4, B31.5, B31.8, B31.9, B31.11).

B31E offers an additional stress check on the strain associated with seismic anchor motion and also changes the allowed stress for the SUS+OCC combination. In my opinion, the B31 code books have not adopted the B31E checks because of these two issues. That's not to imply that the individual books do not accept the physics promoted by B31E. Let's take a look at these two concerns.

Seismic anchor motion - in many instances, the strain caused by independent support motion is more destructive than the inertial loads. The B31 codes throw strain into the flexibility (expansion stress range) calculation. B31E itemizes the seismic strain for separate check here. It's a bit of a strange calculation - the resultant force (both axial and shear) due to this seismic strain is divided by the pipe cross section. That stress must less than yield - not 2/3 yield (as in Sh).

Higher SUS+OCC limit - this gets back to the allowable stress design (ASD) used by B31 and the load resistance factor design (LRFD) used by structural codes. ASCE 7 is often cited as a source for these seismic inertial loads. As a structural code, ASCE 7 uses the LRFD approach. ASCE 7 states that if ASCE loads used for ASD, the ASCE loads can be reduced by 30%, i.e., ASD=0.7(LRFD). But that is only for stress evaluation. Those ASCE loads should be used without reduction if you are after structural response of the piping system - pipe position and restraint load. This implies that there should be two analyses (again, this is my opinion) - one for stress - with the lower seismic load and a second analysis for system response - with the full load.

What I see in the higher allowable SUS+OCC stress is a way to eat your cake and have it too. Use the high load for a proper structural response (no 0.7) and then boost the allowable stress so those loads can also be used for stress evaluation.

Now, your other question - how to isolate those seismic anchor motions (SAM) in a static analysis. First, I must assume that supports are grouped by the deflections of their supporting structures. (If all supports moved in unison, there would be no resulting strain.) If the system is linear, I would run the SAM displacements as their own load case - one for each support group independent of the others. (Any group could be completely out of phase with the others - one structure is moving left while the next one is moving right.) Then these displacement sets should be summed (absolutely?). If your piping model is not linear, I figure an "operating plus SAM" minus "operating" would be the proper approach. But you should also consider "operating minus SAM" minus "operating".

One other B31E note: if using the ASCE 7 seismic g load for piping, B31E states that Rp shall not exceed 3.5. ASCE 7 lists an Rp=12 for typical B31 welded piping. Rp appears in the denominator.

Note that after years 2000, ASCE 7 adjusted the seismic parameters corresponding to a return period as less than 2%-50 years (2475 years return). A factor of 2/3 was considered to scale-down the stress results. I don't think you can apply directly B31E philosophy working with other standards considering 475 years return seismic events (as many other standards consider).