It should be emphasized immediately that waterproofing is the most important aspect for flat roofs and must always be guaranteed. Workmanship and quality control are therefore critical for proper operation and durability over time.
In light of this, it is necessary to clarify that the stratigraphic analysis of a flat roof is far from simple, it requires in-depth knowledge of building physics as well as the ability to master specific software.
The figure above shows a flat roof that has developed serious problems a few years after construction: specifically, a mushroom that has formed in the insulation, penetrated the membrane and entered the house. Tests showed that roof is perfectly watertight, but despite this, the insulation is wet. Who is the blame then?
The problem lies with how hot air moves through the house. During winter, hot air migrates to the outside where partial pressure is lower. Glaser’s verifications states among other things that steam movement occurs only through diffusion; “little steam in a long time”. Perfect air tightness is required for this to happen (Air tightness of timber buildings: a critical and underrated issue!: Part 1 and Part 2), that is the membranes on the hot side need to be properly sealed.
What happens at points of discontinuity? The vapor moves by means of convection or “lots of steam in a short period” and once this excess quantity of steam arrives on the cooler outer layers, it becomes water through condensation. In its liquid form it can no longer pass through the airtightness membrane, instead it builds up and increases the moisture in the insulation and timber elements that may be present (structural beams, etc)
What happens in the summer?
During the summer, the external heat promotes so-called inverse diffusion which pushes hot air from the outside inward, thus allowing for re-evaporation. However, this is prevented if air tightness is achieved by means of avapor barrier(sd≥100m).
It is therefore very important to choose the right membrane, one that has the ability to curb steam during the winter and let it pass when the temperature outside increases. Fortunately, the market has been offering products with these features for some years, now with variable Sd value.
The Sd value which is determined by analysing the stratigraphy, which includes the variable Sd vapor barrier, is not a simple task, it requires software that is capable of simulating both dynamic and non-stationary conditions. It is very important to define the boundary and usage conditions to accurately predict the performance of the roof system over time.
Problems with perfectly water tight roofs due to “unexplained phenomenon” have led some to avoid (prohibit?) flat roofs or believe that software analysis is unreliable or even that flat roofs are only achievable with concrete slabs. We disagree of course.
Determining the sd value of the membrane is only the first step, correct installation of the membrane is just as important if not more. It is therefore strongly advised to use a BlowerDoor test to measure airtightness, through the use of a smoke generator.
These issues are also known in Germany and a few years ago a group of researchers created the “7 golden rules for flat wooden roofs”. The translation of which is shown below.
Conclusions of the speakers at the Congress “Wood Preservation and Building physics” on 10/11 February 2011 in Leipzig on the subject were:
“Unventilated Wooden roofs”
The installation of vapor barriers (Sd ≥ 100 m) on waterproof structures no longer conforms good construction practice. These membranes prevent the activation of “reverse diffusion” mechanisms during the summer which is necessary to dry the winter moisture created by convection of steam which is caused by very small but inevitable air leaks.
7 golden rules for a flat roof. If all 7 conditions are met no specific analysis is required for design of flat roofs.
(in the normal climate according to UNI EN 15026)
The roof has a slope ≥ 3% in “undeformed” configuration or ≥ 2% in “deformed” configuration under load
Dark-colored (radiation absorption ≥ 80%), with no shade
no covering layer (gravel, green roof, terraced lining)
a vapor barrier with sd variable to humidity (inner side)
no uncontrollable cavity on the cold side of the insulating layer and
perfect air tightness and
moisture content of the bearing elements including decking boards (u ≤ 15% ± 3%) or wooden finishes (u ≤ 12% ± 3%) must be checked before the roof is enclosed. These values must be carefully and appropriately documented.
Calculations according to Glaser
Glaser verification is based on the diffusion hypothesis only applies to non-reflective materials, flat roofs without shading with no coating layers. In such cases the drying amount is kept at least 250 g / m² below its capacity as recommended by standards. Note: This test can only evaluate membranes with a constant value of Sd. They must also comply with the rules 5, 6 and 7 above.
Simulating specified temperature and humidity.
Specific simulations according to EN 15026 are required if even one of the above 7 rules are not upheld and therefore the Glaser verification is not permitted.
In this context, the following should be carefully considered:
shading of roof areas caused by the topography of the land, buildings, superstructures (Solar panels, elevator shafts or the like)
waterproofing of upper layers with or without additional insulation
Convection of steam due to less than perfect air-tightness of the building envelope
To this end, the WTA Working Group “Designing the Moisture of Timber Structures” will elaborate calculation and verification rules.
The influence of gravel and green coatings on substrates of unventilated flat roofs with insulation between the joists is still being researched. Currently, flat roofs with cover layers can be verified for safety only if additional insulation (eg sloping insulation) is provided on the top of the coating. The thickness of the insulation required can be found through thermo-hygrometric simulation. Layered roof construction does not require specific checks unless at least 80% of the insulation thickness is placed over the wooden frame.
Richard Adriaans, Herford (D) – Robert Borsch-Laaks, Aquisgrana (D) – Claudia Fülle, Lipsia (Germania) – Daniel Kehl, Biel / Bienne (CH) – Hartwig Künzel e Daniel Zirkelbach, Holzkirchen (D) – Martin Mohrmann, Eutin (D) – Oskar Pankratz, Haidershofen (A) – Ulrich Ruisinger, Dresda (D) Daniel Schmidt, Lauterbach (D) – Hans Schmidt, Bützfleth (D) – Kurt Schwaner, Biberach (D) – Martin Teibinger, Vienna (A) – Stefan Winter, Monaco di Baviera (D), Markus Zumoberhaus, Meggen (CH).
We would like to add an eighth point or the need to continuously monitor the moisture content of timber. Fortunately, there are tools that can do this by means of inserting electrodes in the structural timber frames. For several months now, we are conducting internal tests to verify and better understand moisture content in timber and in a few weeks, we will update you with a specific newsletter.
 The inverse mechanism during the summer correspond to exterior -> interior that means from the hot side to the cold one
 UNI EN 15026:2008 “Prestazione termoigrometrica dei componenti e degli elementi di edificio – Valutazione del trasferimento di umidità mediante una simulazione numerica”.