Why do roofs NOT leak?

Forums › Home Workshop DIY › Why do roofs NOT leak?

1 | 2 | 3 | 4

20515 posts

Uber Geek
+1 received by user: 4795

#3384694 16-Jun-2025 14:19

heavenlywild:

So this poses the question - why do we design roofs this way knowing eventually the water will come through?

I was thinking why don't we put two layers of roofing timber on - sandwiched in the middle a sheet of builder's paper? Wouldn't that make it more waterproof?

I mean nothing is going to be bullet proof but surely there is a better way than today's methods? Wouldn't cost that much more?

cost, plus the current system has worked for many decades without too much of an issue.

mattwnz

20515 posts

Uber Geek
+1 received by user: 4795

#3384695 16-Jun-2025 14:22

heavenlywild:

Batman:

My roof leaks. I've been wondering how to make roofs not leak.

I wonder if it is how we build them? Of course nothing lasts but putting screws in don't make sense from a water proofing perspective.

you can get roofs tray profiles that clip onto hidden clips. So there is no penetration through the sheet. Metdek is one.

pdh

442 posts

Ultimate Geek
+1 received by user: 290

#3384702 16-Jun-2025 14:35

Our roofs leak because we have a benign climate - where it hasn't bothered us enough to change our expectations.

In parts of the world with severe weather - the sort that kills you in hours - roof (and window/door) designs have moved on.

Even many places where the only threats are mould and uncontrolled humidity have learned to do better.

We'll get there - it's was only 30 years ago (1993) that an Auckland Master Builder said I was crazy to want to have/build a warm house.
Every tradie on site (Auckland N Shore) laughed about me using double-glazed windows.
'We don't do that here, north of Taupo, mate.'

Expectations drive change.

richms

29101 posts

Uber Geek
+1 received by user: 10216

Trusted

Lifetime subscriber

#3384705 16-Jun-2025 14:42

also remember that a lot of houses are being used well beyond the 50 years that the systems are designed to last, and its reasonable that they would be failing that long after they were expected to be gone.

Richard rich.ms

pdh

442 posts

Ultimate Geek
+1 received by user: 290

#3384764 16-Jun-2025 15:11

Over the past 10 years, I've built a 2nd new house on Auckland's N shore.
First one was 1993 - the current one dragged on (Covid, Resource & Building Consents, etc).

I speak of 'building' somewhat loosely - as I've had NZ Master Builders do most of it both times.
But (as a MechEng) I have done both the floorplans and had a lot of input into some Architectural details (eg windows).

So I have spent some time looking at best practices around the globe.
Some of that has been consumer-oriented (eg: Grand Designs) - some on various channels aimed at builders.
Sometimes it's possible to push for better-than-the-norm (as I did with DG in 1993).
Sometimes I can only dodge widespread but silly ideas (monolithic cladding systems).

It was fascinating to me (as a lapsed Canadian) to see a good Vancouver builder who completes the waterproof testing of his new houses by literally pumping them up to a low pressure on a Friday afternoon - and doing a leak-down test over a weekend. Don't get the idea that he was unique - it sounded like something that they ragged each other about down at the pub if they'd fxxked up and had to fix something.

That's so far beyond our best-practice that it might as well be the space station...

Another trend is an insulated attic - with the attic space deemed too valuable (for air-con machinery, etc) to be 'cold and damp'.
So a proper impermeable roof has really good insulation applied to its underside - with no concerns about corrosion or humidity.

Not all good ideas are necessarily good in/for NZ, but we could do some stuff better - even less expensively.

mattwnz

20515 posts

Uber Geek
+1 received by user: 4795

#3384766 16-Jun-2025 15:12

Roof spaces also tend to have a lot of air flow to dry out. It is a system that works with Nzs climate

Lenovo

Shop now for Lenovo laptops and other devices (affiliate link).

pdh

442 posts

Ultimate Geek
+1 received by user: 290

#3384768 16-Jun-2025 15:22

>also remember that a lot of houses are being used well beyond the 50 years that the systems are designed to last,

How did we get the expectation that a house (or component) would only last 50 years ?

Leaving aside our earliest wave (1st canoe ?), most settlers came from areas where the expected life of a dwelling was 'forever'.
Just go and travel round the UK (or Holland or Dalmatia)...

I'd say that 50 years was seen as a lower bound - not a target.

The timber house my mum & dad had built in Quebec - 70 years ago - has lasted better than I have.
It's still standing, still looks the same, and should be good indefinitely.

Handle9

11926 posts

Uber Geek
+1 received by user: 9678

Trusted

Lifetime subscriber

#3384769 16-Jun-2025 15:24

heavenlywild:

So this poses the question - why do we design roofs this way knowing eventually the water will come through?

I was thinking why don't we put two layers of roofing timber on - sandwiched in the middle a sheet of builder's paper? Wouldn't that make it more waterproof?

I mean nothing is going to be bullet proof but surely there is a better way than today's methods? Wouldn't cost that much more?

Timber isn't waterproof and is heavy. The more weight you put on the roof the heavier your trusses would be which all adds to the cost. Applying 2 sheets plus paper in the middle would add considerable cost in materials and labour while not working any better.

Handle9

11926 posts

Uber Geek
+1 received by user: 9678

Trusted

Lifetime subscriber

#3384772 16-Jun-2025 15:28

pdh:

It was fascinating to me (as a lapsed Canadian) to see a good Vancouver builder who completes the waterproof testing of his new houses by literally pumping them up to a low pressure on a Friday afternoon - and doing a leak-down test over a weekend. Don't get the idea that he was unique - it sounded like something that they ragged each other about down at the pub if they'd fxxked up and had to fix something.

Pressure tests aren't really meant for waterproofing, generally it's done for thermal reasons. Waterproofing isn't necessarily meant to be sealed, in many cases that is a bad thing as the whole concept of water management is defense in depth. If the house is totally sealed when it leaks the water will have nowhere to go and will sit inside the building and wreak havoc.

mattwnz

20515 posts

Uber Geek
+1 received by user: 4795

#3384773 16-Jun-2025 15:32

pdh:

>also remember that a lot of houses are being used well beyond the 50 years that the systems are designed to last,

How did we get the expectation that a house (or component) would only last 50 years ?

Leaving aside our earliest wave (1st canoe ?), most settlers came from areas where the expected life of a dwelling was 'forever'.
Just go and travel round the UK (or Holland or Dalmatia)...

I'd say that 50 years was seen as a lower bound - not a target.

The timber house my mum & dad had built in Quebec - 70 years ago - has lasted better than I have.
It's still standing, still looks the same, and should be good indefinitely.

Many NZ timber houses are over 100 years too. It is a minimum of 50 years. But a lot of houses doesn't last that , or much beyond because they become unfit for purpose and too expensive to renovate compared to just building new.

pdh

442 posts

Ultimate Geek
+1 received by user: 290

#3384777 16-Jun-2025 15:38

>Roof spaces also tend to have a lot of air flow to dry out. It is a system that works with NZ's climate.

That's classic defeatist thinking...

My Dad (born in London, in Quebec from the age of 17) asked his English brother why all the plumbing pipes were still put on the outside of English houses... The answer was 'so that they could get at them when they froze'. My Dad told that story for years.

If you want to see a wet & humid climate - go and look at the Pacific NW in North America.
Coastal British Columbia & Washington State winters make Auckland look dry (and I've lived here for 43 years).
Our tradition of mould doesn't encourage a belief that what we're doing 'works with NZ's climate'.
We chose tiles / longrun and roofing paper as the best option in 1920 - and haven't stepped up our game since.

Geekzone support

Support Geekzone with one-off or recurring donations Donate via PressPatron.

Handle9

11926 posts

Uber Geek
+1 received by user: 9678

Trusted

Lifetime subscriber

#3384779 16-Jun-2025 15:39

pdh:

>also remember that a lot of houses are being used well beyond the 50 years that the systems are designed to last,

How did we get the expectation that a house (or component) would only last 50 years ?

Leaving aside our earliest wave (1st canoe ?), most settlers came from areas where the expected life of a dwelling was 'forever'.
Just go and travel round the UK (or Holland or Dalmatia)...

I'd say that 50 years was seen as a lower bound - not a target.

The timber house my mum & dad had built in Quebec - 70 years ago - has lasted better than I have.
It's still standing, still looks the same, and should be good indefinitely.

The whole system won't have a design life of 50 years, just the structural elements. The paint system won't last 50 years and generally the roof won't without maintainence.

Maintanence is the element which makes the difference, if it's kept up then the structure will last almost indefintely.

cddt

1970 posts

Uber Geek
+1 received by user: 1905

#3384781 16-Jun-2025 15:42

pdh:

How did we get the expectation that a house (or component) would only last 50 years ?

Happily my house is already over one hundred years old. If looked after well it'll last another hundred years.

My referral links: BigPipe | Mercury

pdh

442 posts

Ultimate Geek
+1 received by user: 290

#3384785 16-Jun-2025 16:09

>Pressure tests aren't really meant for waterproofing, generally it's done for thermal reasons.
>Waterproofing isn't necessarily meant to be sealed, in many cases that is a bad thing as the
>whole concept of water management is defense in depth. If the house is totally sealed when
>it leaks the water will have nowhere to go and will sit inside the building and wreak havoc.

I air-pressure-test my kayak compartments to find water leaks - not heat loss ;-)

We take a very old-school approach to building.
We assume we will fail to keep the water out and very expensively build the structures to allow water leaks to drain out.
Perhaps that started with the European double-brick/cavity wall.

We used to build motor cars and appliances in a similar fashion, until the Japanese showed us that building smarter and better meant we didn't need large remedial-work areas in our factories.

From my (now 40+) year ago experience in Quebec, water ingress was just not an option.
With rapid freeze-thaw cycles any day during the months of March, April & May, you'd blow your structure apart.
Let something get wet in the house's outer envelope (outside the warm/cold boundary) - and that moisture would freeze overnight.
Freezing water to ice expands and blows apart the material or structure.
So there isn't the luxury to 'manage leaked water' - and so houses don't take that approach of letting water in and recovering from it.

I believe we could learn to do the same - and save money.

wellygary

8813 posts

Uber Geek
+1 received by user: 5291

#3384786 16-Jun-2025 16:17

cddt:

Happily my house is already over one hundred years old. If looked after well it'll last another hundred years.

Betcha it has at least 2 or 3 roofs over that time though... Maybe more

1 | 2 | 3 | 4