We had a query last week about the calculation of the U-values for inverted roofs, and the apparently strange behaviour of the rainwater correction factor.

Using the default values for f.x and p (0.04 and 3 mm), the calculation required 2.8 m of insulation to reach a U-value of 0.13 W/m2K. That seemed very strange, so we took a look at the rainwater correction, ∆Ur to try and find out what was going on. That correction is calculated from f.x.p multiplied by the square of the resistance of the insulation (Ri) divided by the resistance of the whole construction (Rt) (you can find the equation in Annex D of BS EN ISO 6946:2007).

For constructions where the insulation is wholly above the waterproof layer the ratio of Ri/Rt is going to be over 0.9, giving a value for ∆Ur of at least 0.1, using the default values. Worse, as the amount of insulation is increased, Ri/Rt tends towards 1.0, and ∆Ur tends to a maximum of 0.12. The final U-value is calculated by adding the correction to the raw U-value, and can therefore never be less than 0.12, no matter how much insulation is added.

The only solution is to use a water shedding membrane system which has a lower value for f.x, preferably with third party certification of that value.