CarlB,
What a warning (to design to?) at the end of the following bolded caveat on page 3 of the document your link points to! (also in same, is it possible "can be" is sort of misleading? -- can pipe withstand indefinite cycling of freezing as it appears might be read by some into this? -- with what length freeze plug and what degree of say endwise confinement? -- what is the specific dimensional and material property/fracture mechanics definition also of "without damaging"?):
"Water can be frozen solid in polyethylene pipe without damaging the pipe, but an ice plug in the pipe will stop flow. Do not apply pressure to a frozen line that has an ice plug. Allow ice plugging to thaw before applying pressure to the line. Severe water hammer (such as from an ice plug stopping suddenly at an obstruction) in a frozen, surface or above grade pipeline can shatter the pipeline and flying fragments can cause death, injury or property damage."
[Frankly, if I had to design a pipeline to provide for specifically for freezing, I'm not sure I don't like better Mr. Halloran's idea of what would appear to be a protective/perhaps even insulating to some extent? casing in an inaccessible area, perhaps even with some sort of double pipe leak monitoring if advisable, or at least casing interstice drain to an area that will not damage if the carrier pipe were to develop a leak, that additionally might allow for more ready repair/replacement of the line if necessary?]