Are you using shingles on a flat roof with no pitch? That won't work. The roof needs have a 3/12 or steeper pitch. For a flat roof or a pitch under 3/12, you need something like a rolled roof.

BTW, a flat roof with one side raised up to provide some pitch is know as a shed roof. Perhaps you could modify the shed to that end.

Here is a link that might be useful: {{gwi:1499258}}

Typically your drip edge is installed, then the underlayment, then shingles.To repair this without pulling shingles off completely, you can, (easier with help), lift the shingles up enough to get some self adhering flashing, (6" wide or wider), and place it over the top flange of the drip edge and over the underlayment. You will have to be careful with the fasteners in the area that are holding the shingles down towards the edge, but if your roof was recently done, the shingles are pliable enough to work with without splitting and you can carefully remove those fasteners that are in the area and replace when you apply the self adhering. Be sure to roll the self adhering down tight with a hand roller upon application. The most used areas of self adhering flashings are used in window, entry door, as well as other wall protrusions, but an asphalt based is decent for roof protrusions,(such as sky lights and vents), and fixes/repairs as well. You could also look into using some ice shield material as well.

Here is a link that might be useful: self adhering flashings

The underlayment being under the drip edge is fine, I got that backwards but the flashing method should work. If you have dimensional or architectual shingles, out here it's common to use ridgecaps up the rakes of the gables. Half the ridge cap runs down over the facia, the other half lapping over the shingles lapped same as the ridge application, (from the bottom of the gable on up,of course). It looks good and keeps winds, (torent in our area), from getting under and lifting the shingles. Might cap off your problem using those as well.

Shingles without overlap will leak.

A starter course of the upper half of 3 tab shingles is placed under the first full course.

Even roll roofing gets a starter course.

Out here, starter shingles are typically used on the leading or eave edge of the roof only, not up the rakes. I've always used them up the rake as well for cosmetic reasons and extra support of the courses overhanging the rake. I believe the op's picture shown as if you are looking at the leading edge on up, and as you stated, doesn't show any starters. I'm assuming the courses have the proper overlap, not shown on the pic. Maybe ,ororke, you could give some more detail on your installation and shingle type.

Sorry, I confused the issue with my poorly presented initial diagram. Here is a better attempt:

{{gwi:1499262}}

As Sierraeast pointed out the diagram shows Â."looking at the leading edge on up," (at a 5/12 angle above the horizon)

I showed two rows of shingles since, with a standard shingle layout, at any cross section there are at least two rows of overlapping shingles. I also showed sierracastÂs self adhesive flashing suggestion (if I understood it correctly).

Shingles are 3 Tab (GAF Timberline Premium Architectural Shingles 30-year)

Roof pitch is 5/12 , so certainly the water on the edge of the shingles runs downward. But because of the additional small pitch towards the structure the water also has a tendency to run inwards (toward the structure) as it runs down the edge of the shingles. So eventually some water gets between the shingles and the underlayment, then eventually on the deck since this is the edge of the underlayment.

The first row at the eave does have the standard "tabless" starter row per manufacturerÂs specs.

A) Sierraeast, I think I understand what you are suggesting with the self-adhesive. The way I understand it, the self adhesive would not necessarily prevent water from getting under the shingles (actually wouldnÂt it make it worse since the thickness of the self adhesive flashing would raise the shingles even more?) but the self adhesive flashing would keep the water from going under the underlayment. Seems though I would still have water between the shingles and the underlayment.

B) I think the ridgecap solution will work, though IÂd have to trim off the shingle overhang (overhang is about 1" - per manufacturer suggestion)

C) Perhaps another solution may be to put some sort of drip edge under the shingle overhang? (though I have never seen this done so there must be a simpler way).

D) I guess I could caulk between the shingles and the drip edge, but IÂm not sure the caulk will hold summer winter without separating. Besides, caulking at that position (either from the top of the roof or a long ladder) is not exactly easy work.

Overall, I am a little bewildered by the fact that since this has been the standard way to construct a roof for a long time(at least before ridgecaps at the rake became more common) most shingled roofs with gables must suffer from this problem? So millions of roofs have this water penetration vulnerability with shingles tilting slightly upwards at the rake?

Perhaps IÂm making a big deal about this when in reality it is not a problem after all?

Something is wrong with the drip edge installation.
It is not thick enough under the shingles to cause what you have described.

Was it kinked and bent and not installed flat?

I think you are right on about the self adhering only making it worse, and while it would redirect the water from getting to the decking, it would mean the underlayment would be responsible instead of the shingle/underlayment combination. You probably dont want to rely on caulking as well as that would be a constant maintenance issue as caulking is typically more of a temporary fix and for exterior is more for cosmetics,(like on trims), than for keeping out water permanently. As Brick mentioned, if the drip isn't defective, could it be fasteners on the sheathing to rafter connect sre sticking up causing the drip to raise? I've seen that happen when the framers didn't send the nails home. How heavy is the gauge of drip edge that you are using?

I am not sure why the drip edge is not completely flat. DonÂt imagine a huge thickness, perhaps only 1/8" but it is enough to give the shingles this undesirable negative tilt (that would be 1/8" slope over the 1" overhang).

- As you say, perhaps the builders (that would be rookie builder orourke) may have not driven the nails to the rafter all the way (not much support to nail this last rafter Âhammering there is flimsy).
- Some of the nails used to secure the underlayment may also be adding thickness under the drip edge.
- Driving a nail to secure the shingle right past the drip edge probably compounds the problem.
- Finally the drip edge itself is not totally flat. It seems to be a bit domed. The drip edge got squashed flat where it was nailed but between the nails it retains some of its original dome.
- Also squashing the drip edge with the nail may cause it to buckle some at the edges. As I said, thereÂs not much room for error, a 1/8" does it.

Drip edge metal thickness itself is quite thin, not sure the gauge, standard drip edges that HD carries.

If I had to do it again, I would have perhaps added a, say 2-3" strip of shingle material along the rake about 2" in from the edge. That would have added thickness and would have caused the newly installed pliable shingles to bend downwards (summer heat helps). Actually, when I had initially installed the (then pliable) shingles, I had hoped that the 1" standard overhang would have been heavy enough to make the edge of the shingle bend downwards and compensate for this but apparently I was wrong, the shingles were too stiff (I think it was winter) - and are even stiffer now 2 years later.

In any case, perhaps I will go up there someday when it starts raining to see how the water is actually behaving. Then perhaps I will take sierracastÂs suggestion, trim the overhang and install ridgecap shingles at the rake. Or take some roof mud/tar and apply it under the shingles at their very edge to create some sort of drip edge on the underside of the shingles.

"DonÂt imagine a huge thickness, perhaps only 1/8" but it is enough to give the shingles this undesirable negative tilt"

Something is seriously wrong if the drip edge is 1/8 inch thick under the shingles.

Drip edge for 3-tab shingles is not nearly that thick.

It sounds like the wrong type of drip edge.
Those dimensions are more like a built up roof.

Sorry, with the term thickness I meant if (nails sticking out + drip edge deformation + nails securing underlayment + drip edge crown ) build up to a total of more than 1/8"Â
The drip edge metal itself is very thin.

Actually hereÂs a picture that I took when I was shingling the roof 2 years ago.

One more thing I remember now: I think I was applying a bead of caulk (tar) between the drip edge and the underlayment as well as a bead of caulk between the underlayment and the shingles at the rake edge. While that may add some resistance to water penetration it may have actually added thickness and made the edge of the shingles lift.

In the picture, you can see the drip edge and you can see the last shingle lifting up a bit (unfortunately that scrap piece of 2x4 is covering up some of the view). The one thing I can see now is that I nailed the drip edge off center (I guess I did not want to put a nail hole close to the edge?) perhaps that allowed the edge of the drip to stay a little higher. If I ever do this egain IÂll make sure I put lots of nails in the center to hold the drip edge down and make sure I drive them well down. Or perhaps IÂd install the last rafter a little lower to give the last 12" of the roof deck an outwards slope?

{{gwi:1499263}}

I think you have hit the roofing nail on the head with the two beads of caulk. Those two beads would give you quite a "ridge". That coupled with the way you have the drip edge nailing held back on the back edge of the drip is more than likely your problem. The caulking was a good thought and intention, but sometimes well intentioned applications do more harm. I say this from experience as I have a tendency to go overkill on applications when it really isn't warranted and ends up being a waste of time and money when it's not necessary in the first place. If possible, maybe you could get under there and flatten down the caulking beads without harming the shingles, but it sounds like your roof has been down awhile, so it might do more harm than good. Another thought would be to lay a 2x4 along the rake over the drip edge area and clamp it down while the weather is still warm to see if that would help lay it down and flatten the caulking, but again, your roof has been on awhile. Might be worth a shot. Live and learn! I'm still vertical and still learning!

I think you are stressing way too much about nothing regarding the rake edge.

I would be considerably more concerned with 2 other things noted in your picture. First off, your nail location appears to be at least one inch to high. If you check your installation instructions, you will find that you should be nailing just above the exposure line. Where you are nailing now, you are creating a "pocket" for water to sit in when it gets blown up. In that "pocket" is a hole for water to seep into your roof deck. Also, by nailing up that high, you are in effect lifting or eliminating some of the weight at the bottom edge of the shingle that is needed to allow it to seal to the seal strip. Big mistake, me thinks.

#2 problem appears to be what is called railroad tracking of your plywood joints, or stacking the sheets. While it wont cause failure, you dont get the diaphragm action of a staggered sheathing, which provides racking strength.

The picture may be misleading, but thats what it looks like to me.

Here is a link that might be useful: Installation instructions.

Sorry, probably too late if it was shingled 2 years ago, Didnt catch that.

GAF 30 yr instructions show the nail line at 5 3/4" to 6 1/4" above the shadow line of the shingle. The shadow line is at 5 3/4 inches, so they want the nail slightly above the shadow line.

In the photo the rake drip edge appears to be installed before the underlayment. This is not the standard method and could lead to the problem you describe especially if then asphalt plastic cement had to be added to seal the underlayment to the drip edge (probably unsucessfully).

macv, you are right, my picture shows that I installed the underlayment over the drip edge at the rake, so actually my diagram and my photo do not agree. And the two beads of caulk probably account for most of the buildup and shingle curl up.

I am curious what instructions you relied upon.

Metal drip edges have a folded curve on 1 side, and a totally flat other side.

The curved side is supposed to be installed over the fascia. The totally flat side goes to the roof. It would be possible to install drip edge backwards, which would cause the problem being described.

Possibly you installed the drip edge incorrectly?? (backwards) Also - you mentioned having some problems nailing. I have noticed that the slightest upward nail - will significantly lift shingles....

Well, IÂve been trying to remember what instructions I relied onÂNot sure if anyone is interested in this but here is what happened:

I did not want to lay down all the underlayment first because it has a tendency to rip if you walk on it. So I installed the drip edge and then was adding underlayment as I was shingling. At some point I probably figured out that "Oopps underlayment should go under drip edge at rake  but it is too late now that the roof is partially shingled - many roofs donÂt even have a drip edge, so itÂs probably no big deal that mine is under the felt paper... I feel rather nervous working on a slope at this height and would rather get done as soon as possible etc."
-----

The drip edge was not installed backwards. The edge that goes on the roof is much wider (4"?) while the edge that goes along the fascia is much narrower (1"?) and has a bent lip to it.