The effect is not limited to the restrained part of the line, but is carried to the results of partially restrained part as well. Typically a partial retrained condition occurs in buried piping between a change in pipeline direction (bend or a tee) and fully restrained point. The max of this 0.2SH is at fully restrained point and 0 at the point of minimum restraint (bend or tee)
Moreover when defined as fully restrained CAESAR will not include longitudinal pressure stress in axial stress calculation. Again given axial stress (assuming compressive) will be overestimated by 0.2SH at the point of minimum restraint and 0 for fully restrained point.

Therefore in my opinion when bourdon effects are active, 0.2SH (tensile) should be added globally to portion of the pipeline defined as filly restrained.

Very often thermal beam bending stresses are added to axial stress to calculate SL and thereby equivalent stress by B31.8 equation 833.4.a/b and B31.4 equation 402.7, an approach that is questionable, but when followed should consider this correction in axial stresses.

Best Regards,

Tanveer Mukhtar