Flat roof insulation thickness — about 150mm PIR for 0.15, 140mm for 0.16

Does the thickness change the condensation risk?

Hardly — in a warm deck the layer order does the work. Run this build-up through U-Monkey’s free BS EN ISO 13788 check at 100mm, 140mm and 210mm of PIR and the result barely moves: condensation under the waterproofing peaks at about 7 g/m² in spring and 3–4 g/m² is left at the end of the year, so with the waterproofing at BS 5250’s figure for bitumen felt, sd 90 m, over a 500-gauge polythene VCL, it fails at every thickness. What moves it is the vapour control, not the insulation: a high-performance VCL clears it, and the check’s How to fix it gives the rating it needs, lower once you enter the membrane’s declared value.

Where thickness does matter is when it is split to save height. Put 60mm of PIR above the deck and 150mm of mineral wool between the joists below it, leave the vapour control layer on the deck, and the check puts heavy condensation on the underside of the deck every winter, more than dries out by summer: a fail, with its own timber check warning about the deck. Move the VCL down to the ceiling and the amount falls to a few grams, but it still does not clear each year, so that fails too. Solve the thickness first, then check the build-up you will actually build.

It is a screening to BS EN ISO 13788 (the Glaser method), honestly labelled as one: it does not model air leakage, rain or moisture stored in materials, so for high-risk or heritage constructions a full hygrothermal simulation (BS EN 15026) is the right next step. Figures above use the app’s defaults — Manchester (Ringway) climate, humidity class 3 (dwellings with high or unknown occupancy), a 1-in-10 design year and, on a roof, the surface taken 2 K below the air for night-sky cooling — so pick your nearest weather station for your own build-up: the result can change from one station to another.