How Is a Leaking Chimney Diagnosed?

Before the Roof: Is It Even a Leak?

The first question is not which part of the chimney. It is whether water is entering the building at all.

Two things produce damp on a chimney breast without any water coming in. Hygroscopic salts from old soot, which draw moisture out of room air and keep plaster permanently damp-seeking. And condensation inside a flue that has been sealed at both ends when a fireplace was removed.

The test is what the damp tracks. Follows rain, particularly from one direction: water. Follows cold weather or humidity, and happens in dry spells: not water.

This costs nothing to establish and it saves a scaffold. It should always be the first question.

Then the Loft

Next, the roof space at the point where the stack passes through, in daylight with the lights off.

What we are looking for is where the staining is and, specifically, the highest point it reaches - because water runs down and never up, so whatever is wrong is at or above that point.

Staining on the uphill side of the stack points at the back gutter. Staining down one side points at that side flashing or the masonry on that face. Staining below the stack points at the apron. Staining on the stack itself inside the loft, rather than on the timbers, points at the masonry or the flue rather than at any of the lead.

That single observation eliminates most of the candidates before anybody gets on the roof.

Then the Roof, Top Down

The order matters, because anything above can put water into anything below and diagnose itself as the lower fault.

1. The pot and the flaunching. Is the pot sound and seated? Is the flaunching cracked? A hairline crack collects water into the top of the stack and produces damp several metres below with everything else in order.

2. The masonry. Depth of the mortar joints on each face, and the condition of the brick faces. Eroded joints and spalled brick on the weather face are the commonest cause of a leak that survives a flashing renewal.

3. The back gutter. Behind the stack, invisible from everywhere else. Is it there, is it deep enough, is it clear? On an exposed hill property this fills with moss and grit surprisingly fast.

4. The side flashings and soakers. Is the lead in its chase, wedged and pointed? Are there soakers under the stepped flashing, or is the flashing doing a two-part job on its own?

5. The apron. Split, sagging, or lifted at the free edge?

And throughout: is any of it a cement fillet, and has any of it been sealed? Both change the answer.

The Macclesfield Question

On the higher ground east of the town, and out towards the hill villages, there is a sixth question that belongs early in the sequence: which direction was the wind in?

Exposure changes what a chimney can cope with. Driving rain from one particular quarter will saturate one face of a stack and leave the other three untouched, and it will get past a lap or an upstand that handles vertical rain perfectly well.

So on an exposed site, a leak that only follows a westerly is telling you which face to inspect before anybody goes up. It is genuinely the most useful thing a customer can report.

The tall stacks on the older silk-era terraces make this more pronounced - a stack standing well proud of a roof on a hillside is in the most aggressive part of the wind field on the whole building.

When It Still Will Not Show Itself

Some chimney leaks are intermittent enough that nothing is wet on the day anyone looks. That is when a water test earns its time.

The method matters: wet one element at a time, starting at the bottom and working up, with somebody watching inside. Start at the top and the water runs over everything below it and tells you nothing.

So: the apron first, then the sides, then the back gutter, then the masonry face, then the top of the stack and the flue - waiting between each. When the damp appears inside, the last element wetted is the answer.

It is slow and it is not glamorous. On a chimney that has already survived two repairs, it is considerably cheaper than a third guess.