Roofs and moisture

I don’t profess to have all the answers, as I have not done original research in this area. However, I would like to inoculate you against the easy solutions that are flogged in public fora.

Let’s start by considering the space in a sloped roof above an insulated ceiling, which I will call the attic. We will further assume that the attic is above a heated space.

Possible regular entry routes for water, be it in liquid or in vapour form, are:

1. water or water vapour entering from outside

2. water vapour transported into the attic via convection

3. water vapour transported into the attic via diffusion

Let’s discuss them one-by-one:

1. I am excluding the option that your roof leaks and lets liquid water in. Should that be your situation, fix that leak, and don't bother reading on.

Water vapour may enter the attic from outside with ventilation events. Particularly during clear nights, the roofing can cool below air temperature due to radiative heat exchange with the night sky, and air that is brought into the attic with ventilation may condense under the roofing. Depending on the materials in the mix, the ability to drain moisture, drying potential based on climate, shading, ventilation and a few other things, moisture may accumulate, drain away or just evaporate again.

2. Air leaks in the ceiling or, particularly with timber frame construction, also air leaks in walls and floors may transport moisture-laden, warm indoor air into the attic driven by buoyancy and/or water vapour pressure differences. This can be large amounts of water vapour. Depending on the amount of water vapour, construction details, and temperature in the attic (influenced by the general climate, but also orientation, colour and shading of the roof, as well as ventilation of the attic), moisture may condense and accumulate over time.

3. Diffusion means that water vapour squeezes through micropores in building materials driven by the higher water vapour pressure of the warmer air below the attic during the colder months of the year. This can reverse when the attic gets warmer, or equally warm than the spaces below, but more moisture-laden. Compared to problem 2, this is typically the lesser evil by a large margin.

To deal with problems 2 and 3, we can utilise control layers: an airtightness and vapour control layer, which may be the same membrane. Ideally, the vapour control layer should curb water vapour rising up during the colder months, but not water vapour diffusing out during the warmer months. For timber-frame construction, we not only need those layers to be gapless in the ceiling, but also in the walls, and even connecting lightweight floors, as air and water vapour can travel in all directions when pressurised.

Problem 1 is more difficult to grapple with. We could simply exclude ventilation from the attic, but as ventilating can be quite useful for removing moisture and excess heat, it shouldn’t be discarded lightly. We need to carefully look at the climate, the orientation of the building, prevailing winds, underlay, roofing materials, colours, shading, and the remaining moisture load from problems 2 and 3 to properly assess the best way forward. A vapour-open membrane on the outside of the insulation, which is preferable in a ventilated cavity to prevent wind-washing anyway, may help with keeping the insulation dry, and make evaporation of any water collected during clear nights easier to evaporate during the day.

Lastly, are all of these problems avoided simply by insulating between the rafters under the roofing instead of insulation in the ceiling? Nothing changes for convection and diffusion with this approach, although it will be more difficult to create a gapless airtightness layer at the wall-roof joint with a truss roof; other roof constructions may be easier but are not very common in Aotearoa. There is a large body of literature attesting to the failures of warm roofs overseas. The only problem that can be lessened with this approach is the problem created by radiative heat exchange with the clear night sky. While the roofing may still get colder than air, and any air circulating under the roofing may still bring moisture with it that condenses there, if your roof has a proper underlay that drains freely, this water will not create damage. Now, we also have examples of warm roofs working well - I am just saying, it isn't guaranteed. The downside of insulating between the rafters, apart from the added difficulty with creating gapless airtightness for truss roofs, is that you have to condition an added volume of air that you are not using, which unnecessarily increases the heating requirement.

As per the preface: I don’t have all the answers, and nothing that I am saying here pre-empts properly assessing the situation at hand. But I hope the above enables you to ask the right questions of the people who profess to know what to do.