(it appears on page 4 of the 2002 EPA document that 15% of the “more than 100” actual pipe permeation incidents reported in that particular case study were with pvc pipe. Below that it appears reported there were also permeations through pvc, pe,and polypropylene services (it appears at not grossly different incident rates of ~2.2-16.5 per million out there, with th e2.2 being for pvc)).
I also understand there has been some research done at Iowa State University not long ago with regard to this subject (associated with an AWWARF grant?) that was reported at the AWWA ACE 06 in San Antonio, Texas.