Why Do Chimney and Valley Leaks Keep Returning?

Why These Two Specifically

Most of the roof sheds water. Chimneys and valleys concentrate it, and that changes how much a small defect matters.

A valley takes the run-off from two slopes and funnels it into a channel a few hundred millimetres wide. A chimney interrupts the flow of water down a slope and makes it go round, so everything above it is directed into a back gutter and out along both sides.

Both are also invisible. You cannot see a valley from the ground on most houses, and you cannot see the back gutter behind a stack from anywhere except the roof. Which means neither gets inspected unless somebody goes looking.

Put those together and you have the two places where a repair is most likely to address the wrong thing.

Reason One: The Wrong Element Was Repaired

A chimney has five candidates, not one: the flaunching at the top, the pointing, the brickwork itself, the four pieces of lead, and the flue.

Flashing is the visible, understandable one and it is what most people are quoted for. But if the water is coming through cracked flaunching, or through eroded joints on the weather face, or down an unlined flue, new lead changes nothing. The leak continues and the homeowner concludes roofing is a racket.

A valley has fewer candidates but the same trap. If the cause is debris damming the channel, relining it fixes nothing that was broken and the dam rebuilds itself by the following autumn.

Reason Two: The Repair Was Sealant

Mastic over a failed junction is the single most common bad repair we find on both.

It sheds water for a season, then shrinks and cracks. Because it is bonded at the edges it now holds water against the defect rather than letting it drain, so the leak comes back worse than it was.

Lead junctions work by shape and gravity - laps, upstands and free channels sized for the volume. There is no product you can apply over a failed junction that turns it back into a working one. Anything that has been sealed needs the sealant removing and the detail forming properly, which is more work than if nobody had touched it.

Reason Three: There Are Two Sources

If two defects put water into the same place and only one is fixed, the leak appears to have survived the repair.

This happens at chimneys constantly, because the same stack can have a lifted side flashing and eroded pointing on the weather face. Fix the lead, and the masonry keeps letting water in. From the ground that looks like incompetence; on the roof it is arithmetic.

The defence against it is checking everything uphill of the entry point rather than stopping at the first plausible fault - which takes longer and is the reason a proper inspection is worth what it costs.

Reason Four: Volume, Not a Hole

This one is specific to valleys and to back gutters, and it explains the "only in heavy rain" leak.

A valley has to have enough clear channel between the cut tiles either side to carry the flow from both slopes. Cut them too tight - a workmanship fault rather than wear - and the valley copes with ordinary rain and overflows sideways in a downpour, entering the roof several courses above where anyone would look.

A back gutter behind a wide chimney on a shallow pitch has the same problem. Formed too shallow, or silted with grit and moss, and it overflows into the roof every time it rains hard.

Neither of those is a hole, so neither shows up on a dry-day inspection. Both are cured by giving the water somewhere to go rather than by patching anything.

What to Ask For on the Second Attempt

A water test. Wetting one small area at a time, from the bottom of the slope upwards, with somebody watching inside. It is slow and unglamorous and it produces an answer instead of a theory. If a leak has already survived two repairs, it is the right thing to do rather than the third guess.

A written finding that names the defect and its location, not just the work. That is what lets you tell whether the next repair is addressing something different from the last one.

And on a chimney specifically, confirmation that the flaunching, the pointing and the brickwork were looked at as well as the lead. If nobody can tell you what condition those were in, they did not check.