I agree that predicting heave from the oedometer test is a little tricky. Corrections should be made for sample disturbance, etc. Determining the proper swelling index and corrected swelling pressure can have substantial effects on predictions of total heave. I used to think that my heave predictions were always on the high side, but after a main water line broke and saturated the clay subgrade of a local building, I now wonder if my heave prediction may be a little on the low side. I found a pretty good research paper that parallels my findings in northern Alberta. "The prediction of total heave of a slab-on-grade floor on Regina clay" - R.T. Yoshida, Canadian Geotechnical Journal, 20: 69-81, 1983". I am a little partial to Del's work, he's a Saskatchewan farm boy such as myself.