For predicting burst pressures you could use CorLas, modified NG-18, PAFFC, even KAPA (has option for crack like) or API 579 & BS 7910. CorLas seemingly shows the best correlation in burst pressure tests. You would need to know the toughnes values of the pipe for increased accuracy. Then as per stanweld you would need to get a proper length and depth of the crack, for SCC I would recommend to use maximum interlinked length.