Chimney problems
A new client has hired us to solve a chronic chimney leak problem. It is a unique convergence of roof angles and chimney. they have had a new crown installed in the recent past and it is in good shape. Their flues are terra cotta and are not capped. They said they had it pointed, but the photos show some need here. The only thing that has not been reworked is the flashing. I think removing the rubber and replacing with soldered copper pans, steps, and regletted counters may be in order. I am wondering if the cricket configuration, given the steep roof and large volume of water on the cricket side, should be altered. here is a link to the images:
http://www.flickr.com/photos/65299059@N05/sets/72157628417670961/
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Robert,
There are lots of potential problems here.
It's impossible to say how the water is getting in without a site visit. The water could be entering the brickwork via the cap.
Some large brick chimneys like this one can't be made waterproof without installing through-chimney flashing -- a type of copper flashing that requires all the brickwork to be taken down to roof level. Then the through-chimney flashing is installed and the chimney top is rebuilt.
One thing is for sure: this chimney does not have proper counterflashing. Counterflashing needs to be set into the mortar joints. One side of this chimney has metal flashing held against he bricks with screws and then smeared with silicone caulk -- a horrible job. The other side is even worse -- the counterflashing is slathered with black roofing cement, and then the black roofing cement is slathered with silicone caulk.
All of this bad flashing has to come out. You need to get a real roofer and a real mason on this job.
For more on through-chimney flashing, see an article by Jerry Carrier, "Keeping Water Out of Brick Veneer," from the November 1999 issue of the Journal of Light Construction:
http://www.jlconline.com/cgi-bin/jlconline.storefront/4ee8d99902850e0a27170a323cb4065e/Product/View/9911keep
uuuuugly
but where is the water coming in?
I had to cap an unused flue because it would leak out 20 feet down in a bathroom
[edit]
Wonder how long it took Martin to clean the coffee off of his monitor after he saw that flashing
Keith,
I think some of the coffee came through my nose. I'm OK, though. I've seen it all before.
Robert,
In some cases, you can narrow down the source of the water leak by getting a garden hose up on the roof and spraying sections of the chimney. You need someone indoors with a walkie-talkie or a cell phone.
The water has shown up in a room on the second floor. The block has been exposed there and gets very wet during prolonged heavy rains. it is on the cricket side of the chimney. The crown of the chimney is in very good condition but the 2 clay flues are open, no spark arresters or caps. The house is about 25 years old.
One concern is the design of the cricket given the steep pitched large roof. The other, after reading Martin's comment, is if we completely re-do the roof/chimney intersection with properly installed copper, will the problem persist.
Looking back over my career, I've solved maybe 30 of these, but there was one about 15 years ago where we re-did the flashings and moisture still occurred on the masonry in the attic. We then air-sealed and insulated that attic floor to try to isolate the interior heated air from the cold masonry in case it was condensation. It was a 80 yr old home with a very ornate brick chimney. We confidently repainted the stained ceiling only to have staining return a few months later. Problem not solved. I wonder if this moisture bypassed the flashings and entered through capillary action through brick an mortar like Martin described.
Robert,
Water can be soaking the cap (if the cap is concrete or mortar); water can be soaking the bricks and mortar, and entering right through the bricks and mortar; water can be getting behind the gunked-up counterflashing; or water can be getting through flaws in the cricket.
Eliminate some of the possible entry points, if possible, with a garden hose. As I said, in some cases you need to install copper flashing right through the bricks, with the copper flashing touching the flue.
The moisture coming through the mortar (or lack thereof) is a red flag. I bet heat is escaping and causing the condensation seen in the photo. May be dripping on the inside.
I'm not seeing a cricket at all.
Dave,
Except for the fact that Robert tells us that the interior "gets very wet during prolonged heavy rains."
There is a cricket of sorts viewable at flickr link. Im not viewing this as a condensation problem. The wet spot on masonry could be from morning dew or excess moisture within the masonry finding its way out at the mortar void.
Does the abundant presence of efflorescence tell us anything? It is present on the side where the leak is occurring below in the house.
Also, if everything is addressed except the through wall flashing, could the cap and bricks be treated with a clear sealant to minimize absorption or would this potentially cause freeze/thaw vulnerability?
Robert,
This is what Jerry Carrier had to say about coatings in his 1999 JLC article:
"Some people install brick veneer without flashings, thinking they can always fall back on the application of a coating if they have a water penetration problem. However, a coating can’t bridge bond breaks or hairline cracks. If you use a coating on bricks, remember that you are taking a chance that the coating won’t work.
"In general, clear coatings sometimes work to prevent moisture entry. However, coatings can also create problems. If the pores of the bricks are coated, and water gets into the bricks from the back side, where the bricks aren’t coated, then the water can’t get out. When the saturated bricks freeze, the expansion of trapped moisture within the brick can cause severe deterioration.
"Some coatings are breathable, which means that the coating allows moisture to exit the wall as vapor. But remember – if a coating is applied, the pore structure of the brick has been altered. The pore structure directly affects the durability of the brick. There must be enough pore space in the brick to allow any absorbed moisture to freeze and expand. If the pore structure is altered and the expansion room is no longer available, the brick may deteriorate.
"If you decide to try a coating, choose a breathable coating. While some types of silanes and siloxanes have proven to reduce moisture penetration without reducing the durability of the masonry, the siloxanes appear to be more successful with brick.
"The siloxanes are fairly expensive. You get what you pay for when it comes to coatings. If it costs $5 a can, you can do without it. In any case, properly installed brickwork should not require coatings to prevent moisture penetration."
VERY HELPFUL. Thanks so much for your expertise. I've used Radon Seal in the past and the same company makes a Siloxane sealer as described above. That may be part of the solution depending on the water test and the owner's desire to rebuild or not.
Yes, per Robert's comment #11, that efflorescence is definitely an indication that rain is entering the stack at the top and looking for ways out, probably also to the interior. This needs to be fixed. But that should not be the only concern. Given the generally poor detailing evident in the photos the exterior surface of the home with its various elements of brick, stucco and applied boards needs an overall inspection and overhaul. Attention to rainwater management should obviously be paramount. A couple of glaring examples of flawed guttering apparent in the photos:
• The arrangement in picture #2 is obviously unsatisfactory, the downspout from the upper roof with its triple bend is obviously leaking/overflowing and the adjacent veneer is clearly saturated on a regular basis. And guess what - the veneer is here directly connected to the stack brickwork and hence the interior of the house, with obvious potential for concealed damage to structural framing. This upper section of roof needs its own downspout directly to ground, preferably away from that vulnerable corner.
• A later photo shows mold staining which indicates another area of saturated veneer at a stepped gable. I'm not sure what the complete answer is here, the flashing of the projecting piece of roof to the adjoining veneer probably needs attention (is there any?) and the short piece of tin gutter needs to be replaced with an arrangement with sufficient capacity to absorb peak rainfall and direct it reliably into the pipe, perhaps a decorative hopper.
And boy, overall, is this the textbook example of how NOT to design a roof, as per Martin's recent excellent blog post. This is a triumph of style over substance. If you're going to design a roof this complex you have to pay far more attention to details than is exhibited here, and use better quality materials as well. Worth noting that the style originals would have been constructed with slate roofs and robust, properly coursed-in lead flashings and counterflashings.
James,
Excellent points and good observations. I agree.