I found this:

"The method proposed by Coade, the software developer, in their April 1991 article in “Mechanical Engineering News” is adopted in this analysis. The method suggested by Coade for determining the allowable stresses for plastic pipe is based on the allowable span given by manufacturer. A straight pipe, filled with water, sitting on rest supports spaced at the maximum span stated by the manufacturers catalogue is analysed using Caesar II and the largest code stress and largest bending stress at the support location are taken as the limits to the bending and code stresses that should be allowed in the operating system. However, the allowable stresses for piping so obtained appear to be very low simply because the pipe allowable span table is too conservative."

I hope someone can clarify this more.