I've heard of a few very people getting a Ph.D. in 2 1/2 years. . . but they also had a master's degree first. The norm in my field is 4-5 if you already have a master's (usually 1 1/2-3 years) and 5-7 without one. If things perpetually keep going wrong or you have some other issues to deal with, that 4-5 I mentioned if you already have a master's can definitely be 8-10. . . I've known a few people over the years that's happened to. No, it had nothing to do with their work ethic in any case -- it was a Murphy's Law situation.
If you're talking about going straight from a bachelor's to a Ph.D., you have to factor in that you have a bunch of graduate-level classes you'll need to take -- it's probably going to be at least 60-70 hours if it's like anywhere I've been. Full-time is generally 9 hours in grad school, and I've never really seen anyone take more than 12 hours -- 9, really, is often plenty of work given than you're probably also trying to do some research and might well have some teaching duties as well to manage. You're probably doing the same research project the entire time that way as opposed to someone doing both a master's and a Ph.D. project, but the course requirements are still the same. . . you can bring in courses from a master's for a Ph.D., usually, so all isn't quite lost if you go that route.