When singing opera or other high-voice songs, like this one here, the accent must change a little so that people can understand better. For example, the /r/ sound is so soft that it is very difficult to hear from a distance without a microphone, so this English sound is traditionally substituted for a rolling R like the one you have in Scottish, Spanish or Italian in the word "rosa". This rolling R can also be heard in classic plays (for instance, Shakespeare's) for the same reason (the microphone is a very modern invention).

Another shocking thing here is the way they spell the words "situation" and "torrential": situa-cion / torren-cial. This is because they are dividing up the words to match the musical pattern, and if they wrote situa-tion, then the pronunciation of the second part would be /tjən/ with a T instead of a S (or SH). So writing a C, they avoid misunderstanding.

In this song they use the old fasioned pronunciation of the word:  /sɪtjʊeɪsjən/ instead of the modern pronunciation: /sɪtʃʊeɪʃən/. The present pronunciation derives from the simple rule of assimilation that says that T+Y=CH (/t+j=tʃ/) and S+Y=SH (/s+j=ʃ/).