BETPM · Building Envelope Thermal Performance Model

Outlining the members lets the insulation read straight through them, so where the blanket starts and stops is visible even where a chord or a batten crosses it.

A trussed roof has a horizontal ceiling with a roof space over it: the bottom chord carries the ceiling, the top chord carries the covering, and the space between them is a volume — ISO 6946 §5.4.1 Table 3 gives its resistance directly.

Roof battens run horizontally, across the truss, so the section cuts through them. Counter-battens run up the slope and are seen along their length.

Struts are the same section as the chords by default — change this only if the truss really is built with a lighter web.

Choose None where a board fixed over the top chord doubles as the ceiling, or where the rafters are left exposed to the room. There is then no lining and nothing to hang battens from, so the whole ceiling group drops away and the room sees the underside of the structure.

Steel top-hat battens under timber trusses is a common combination, and the batten’s own λ is what bridges the blanket — not the truss’s.

This is a clearance, not a thermal input. It sets the ventilation path the roof has to keep, it is what the section is drawn to, and it is the depth the roof-space cavity is quoted at — but ISO 6946 Annex B holds h_a on its floor for any cavity deeper than about 13 mm, so making the roof space deeper does not raise R. The derivation says so on the row where the depth is printed.

Ventilated is the default, and it is what a tiled or sheeted roof actually is. Under tiles or profiled sheeting this zone is open at the eaves, at the ridge and through every lap, and ISO 6946 §5.3.3 then disregards it and everything outboard, substituting a still-air Rse at the sarking — which the analytic route now does. The 3-D ASSEMBLY route does not yet: it still adds the roof plane explicitly, so the two disagree about whether the covering counts, and shipping that inconsistency would be worse than shipping the conservative cavity. Select it to see the ISO treatment; the default stays sealed until both routes agree.

The space between the battens, above the sarking and under the covering. It is a DIFFERENT cavity from the roof space below, with its own ventilation, and under tiles or profiled sheeting it is open at the eaves, at the ridge and through every lap. Treating it as a sealed cavity credits it with resistance it does not have — and it is what makes a batten appear to IMPROVE the roof, because a sealed 38 mm air gap is a worse insulator than the timber crossing it. Ventilated, ISO 6946 §5.3.3 disregards it and everything outboard of it, and an external surface resistance is taken at the sarking.

How to get this number off a drawing. Free area is the total clear opening into the roof space, in mm², divided by the ceiling area it serves, in m². Nothing on a drawing is labelled that, so it is worked back from the eaves detail and the span. A continuous gap of g mm at both eaves gives 2 × g × 1000 mm² for every metre of eaves run; a vented ridge adds about 5000 mm²/m, BS 5250 treating high-level ventilation as a continuous 5 mm gap; and one metre of run serves span m² of ceiling. So free area = (2g×1000 + 5000) ÷ span. Measure the gap NET. Insect mesh, a comb filler and a bird guard each take back roughly half of it, and it is the clear opening that ventilates, not the gap it sits in. On a mono-pitch there is one eaves, not two — enter half the gap, or put the high-level opening on the ridge line. The span is what decides the class, which is why it is asked for: a 10 mm eaves gap on an 8 m span is 2500 mm²/m² and well ventilated, while the same detail on a 20 m span is 1000 and only slightly ventilated. ISO 6946 §5.3.1–5.3.3 then reads the class off it — above 1500 well ventilated, over 500 up to 1500 slightly, 500 or less unventilated — and the class, not the number, is what changes R.

The type decides whether it can be compressed at all. A rigid board spans purlin to purlin without sagging and without crushing under the fixings, so laid over the purlins it is a genuinely continuous layer with the structure entirely inboard of it — SANS 10400-XA Figure F.1. A fibrous blanket cannot do that.

On the ceiling board assumes the roll runs ACROSS the trusses, so it rides over the members and part of the depth stays continuous. Between board and chord assumes it runs along the bay, so it is limited by the batten depth. Roll direction is not a separate control — it is what the position means.

A full-depth spacer eliminates compression — it does not eliminate bridging. It sits on the purlin and runs with it, so it crosses the blanket on the purlin gauge over its own width. Figure F.3’s continuous solid spacer bridges its full width; Figure F.2’s mechanical spacer is a bracket, so declare a much smaller width for it.

Applies with the insulation over the top chord. SANS 10400-XA 5.6.3 is normative: a roof assembly with flexible bulk insulation between the cladding underside and the top of its supporting member shall have a spacer system, the depth of the spacer equal to the material thickness required, to eliminate compression. So the first option is the requirement and the other two are non-compliant configurations, shown because they are what gets built. Figures F.2 and F.3 illustrate a mechanical and a continuous solid spacer.

Insulating at the ceiling and on the roofline is common, but on a ventilated roof the two are not simply additive: ISO 6946 §5.3.3 disregards a well-ventilated air layer and every layer outboard of it, so a roof-line layer above a well-ventilated roof space earns nothing at all. The app applies that rule rather than adding the two together.

A foil has no R of its own — it changes the radiative conductance of the cavity its low-ε face looks into, so without an air space it earns nothing (SANS Annex F.4). Direction matters more than the product: the same foil is worth roughly three times more downwards than upwards. Dust ticked selects Table 9's 0,9 outer column, unticked selects 0,2 (Table 9 NOTE 3).

Annex D is normative: U shall be corrected for air gaps and for fasteners that penetrate an insulation layer. ΔU_f = α·λ_f·n_f·A_f with α = 5 m⁻¹ for a roof fixing. Nothing under λ 1 W/(m·K) is corrected at all, so a plastic-sleeved or timber fixing scores zero. The procedure does not apply where both ends of the fastener touch metal sheets — that case needs ISO 10211.

Printed at the head of the calculation. Nothing here changes a number; it is what makes the sheet identifiable six months from now.

Density and specific heat decide how much heat a layer can STORE. They change nothing about R — a stored joule is not a resisted one — but they are the whole of the dynamic behaviour: the decrement factor and the time lag come from these numbers and nothing else.

These feed the decrement factor and time lag, which are computed from the 3-D solve — not from R. Nothing here moves until you solve. Solve in 3-D now

ρ in kg/m³, c in J/kg·K. The roof-space height is here rather than under ROOF because it is the only thing it affects: a taller attic is worth nothing at all to the R-value, and about a quarter of an hour of extra lag between 0,3 m and 4 m, because air has mass.