A SIF of 2.0 based on no data!!! Wow thats amazing!!!! You don't know what the D/T ratios are or what the angle of the Wye is or anything else.

Seeing as how we are picking numbers from thin air why don't you calculate the SIF for a pipe to pipe intersection and double it... That seems "Conservative" to me.

The highest qualty answer which is the best solution is to derive an SIF via skilled competent use of an FEA program. FEPIPE is recognized as one such program used widely for this type of issue.

Now let me go consult my OUIJA board to see if the system I am looking at complies with the code stresses or not.