Not seeing the details and relying only on your description, let me suggest that if it isn't a "black and white" case then the correct action is an educational one, not a disciplinary one. The paper should be corrected and the result evaluated after discussing the issue with the student.
But a clear case of plagiarism, when the "stakes are not minor" would be to reject the paper altogether and let the consequences flow from that depending on other policies. It might result in expulsion or it might require the student to start over, but that is a separate issue.
Another recent question here is similar. What should be done about "suspected but not proved" cheating. My answer was to ignore it. Without proof, a disciplinary action is unjust. (Note: ignore the potential "infraction". The paper itself in your situation probably can't stand without correction.)
For someone who is not a student (i.e. a professor or researcher) the penalties for plagiarism can be quite severe, because professionals should know the rules and have been expected to practice them in all situations. Student's, however are not yet professionals, so their misdeeds need to be treated in an educational context. That doesn't mean ignored, but you need proof of intent to "bring down the hammer".
Students can't be expected to "just know" these things, nor to have practiced them enough to avoid sloppiness.
If you handle it correctly, as an opportunity to educate, you might just save someone from ruination, now or later. The paper almost certainly needs a change if it isn't rejected, of course.