When performing intervals or repeats aimed at raising LT, my standard recommendation is to either 1) make each effort at least 15 min in duration, or 2) keep the rest period between them very short (i.e., 1 min, or ideally even less), so as to trick the body into thinking it is more of a continuous effort. That recommendation is based upon two things:
1) my own experience in training myself and others using intervals of 5 to 20 min duration, which indicates that longer efforts are more effective, and
2) the observation in the scientific literature that short intervals generally do not result in an increase in the respiratory capacity of type I muscle fibers, even though such fibers clearly are recruited during such activity, e.g.,:
http://www.ncbi.nlm.nih.gov/entrez/...st_uids=134623&query_hl=8&itool=pubmed_docsum
Based on the above, my working hypothesis is that the signal for mitochondrial biogenesis is related to both the energy "deficit" that is created
and the time for which it is maintained, akin to the notion that the tension-time integral is an important determinant of muscle fatigue. This would explain why, at least in my experience, shorter efforts aren't as effective as longer efforts in raising LT, even if the intensity and total volume of work are held constant.
EDIT: I might add that John Verheul has independently arrived at the same conclusion re. point #1, i.e., intervals of 12 min or longer seem to work better for him/his clients than shorter efforts.
