On Fri, 04 Mar 2005 15:14:37 GMT, "Travis Jordan"
scribbled this interesting note:

John Willis wrote:
I've seen this comment posted several times and I still fail to
understand how one comes to this conclusion. The surface exposed to
the weather should turn the water. No exception. Any penetration of
water through the shingles is unacceptable as if it happens once, it
will happen again and again and that is called a leak. On rare
occasion a unique combination of events may cause water to penetrate a
roof in such a way that may not ever happen again. When and if that
happens, then and only then does the quality of the felt come into
play. If water happens to penetrate the shingles in this manner, a
good felt job will indeed carry the water off and cause no interior
damage, but if it is a recurring penetration, that felt will quickly
fail (or the legs of the fasteners penetrating the felt will rust to
nothing, leaving the holes they made in the felt-a water penetration
point) and then the problem shows up inside the dwelling.

It is true that the exposed roof surface (ie. shingles, tile, metal,
whatever) will deflect the majority of the water. But wind blown rain
will get under this surface, and that is where the underlayment comes
into play.


In which case it would constitute the secondary waterproofing as it is
the backup, hopefully failsafe, protection that is just-in-case the
occasional bit of water might penetrate the primary waterproofing
which is the first level the weather conditions will contact. At least
that is how I've always thought of it.


--
John Willis
(Remove the Primes before e-mailing me)