For rendered ETICS facades, specify waffled (W) or milled-and-grooved (MG) XPS boards; smooth-skin boards such as XPS 150, XPS 200, XPS 300, XPS 400, XPS 500, and XPS 700 are explicitly excluded from direct plaster and render systems in current PNP product documentation [S1].
All three surface variants share the same declared thermal conductivity, λD 0.034–0.035 W/(m·K) depending on thickness, and comply with EN 13164:2012+A1:2015, so the choice is driven by adhesion mechanics and water-absorption class rather than by insulation value [S3]. The decision matters because XPS is increasingly used as the rigid-foam layer behind thin-coat renders on residential and commercial facades, where bond failure translates directly into hollow spots, cracking, and rework.
What the three surface types actually are
Smooth boards leave the extruded skin intact, producing a dense, low-porosity closed-cell face with long-term water absorption WL(T)0.7 (≤ 0.7 Vol.-%) and no mechanical key for adhesives [S3]. Waffled boards (W, marketed as X-Grip on the 150 kPa grade) are embossed with a regular waffle pattern during extrusion that interlocks with base-coat compounds while preserving the WL(T)0.7 water-absorption rating [S3]. Milled-and-grooved boards (MG) are mechanically milled after extrusion to remove the skin entirely and cut grooves into the foam, which raises the declared long-term water absorption to WL(T)3 (≤ 3.0 Vol.-%) but maximises mechanical key for renders and tile adhesives [S1].
Compressive strength grades 150, 200, 300, 500, and 700 kPa are available across the surface types; XPS 700 (700 kPa) and XPS 500 (500 kPa) are typically reserved for road construction, inverted roofs, and heavily loaded podium decks where a bonded finish is not in contact with the board face [S2]. For a worked example of how a non-render-grade material spec collides with a bonded finish layer, see how spec mismatches cascade into rework in this cladding-failure breakdown.
Selection criteria: where each surface wins
Choose a smooth-skin board when no bonded finish coat is applied to the insulation face: under floor screeds, in foundations, in cavities, behind membranes in inverted roofs, and as compressible fillers in expansion joints [S3]. Choose W or MG whenever render, plaster, tile adhesive, or any cementitious compound is applied directly to the board face, as in ETICS facade systems, plinth insulation with an applied finish, and tile-adhesive systems [S1].
Typical starting points for standard residential and commercial ETICS are XPS 150 W X-Grip and XPS 150 MG at 100–150 kPa; for plinths, accessible areas, and heavy-tile finishes the 200 or 300 kPa variants are the conservative step up [S1]. A smooth board can be harder to bond unless the system is explicitly designed for it, while a factory-textured or grooved surface makes adhesion more forgiving on real sites where dust, weather, and open time vary [S5].
Surface options compared on decision criteria
On the four criteria that drive specification, the surfaces line up as follows. Adhesion to render: smooth = not suitable (no mechanical key), W = suitable (waffle pattern interlocks), MG = suitable (milled face plus grooves) [S1]. Long-term water absorption WL(T): smooth = ≤ 0.7 Vol.-%, W = ≤ 0.7 Vol.-%, MG = ≤ 3.0 Vol.-% [S3]. Thermal conductivity: identical at λD 0.034–0.035 W/(m·K) for all three, because the surface treatment does not change the bulk foam [S3]. Cost: smooth or with-skin surface is the cheapest to produce, with W and MG adding embossing or post-extrusion milling steps [S6].
For most European ETICS facades the trade-off resolves to: MG where the highest pull-off strength and tile-adhesive compatibility are required (and the higher WL(T)3 absorption is accepted because the render layer protects the face from direct wetting); W where the WL(T)0.7 rating is preferred for plinths and ground-contact-adjacent zones; smooth only when the board is fully isolated from any bonded finish [S1]. PNP product data explicitly lists XPS 150 W X-Grip, XPS 200 W, XPS 300 W, XPS 150 MG, XPS 200 MG, and XPS 300 MG as the variants cleared for ETICS, plaster, render, and tile-adhesive use [S1].
Failure modes when the wrong surface is specified
Bonding-related failures, not the XPS itself, are the dominant exterior-wall failure pattern: hollow drum sounds on tap-test, local bulging, and edge lift after temperature swing or wind load are the typical signatures [S5]. The root causes track back to dusty or weak substrate, adhesive not matched to XPS, incorrect mix, insufficient contact area, installation on a wet or contaminated wall, and missing mechanical fixings in high-wind zones, all of which are amplified when the board face offers no mechanical key [S5].
On the moisture side, water rarely moves through the board; it enters through board joints, window perimeters, balcony slab edges, parapets, pipe penetrations, and failed sealant transitions, then shows up as damp staining, efflorescence, condensation at interior corners, or mould growth [S5]. The W and MG surfaces do not solve these detailing failures, but they do prevent the render layer itself from debonding and exposing the board to direct wetting, which is the precursor to most of the moisture problems. For ETICS in particular, a smooth board with render applied directly is, per the manufacturer, excluded from the system documentation entirely, so any pull-off failure on a smooth-skinned rendered facade is a spec error, not a workmanship error [S1].
System build-up and the role of the surface
The standard ETICS plaster system layer sequence on a rendered XPS facade runs: wall substrate, adhesive compound, PNP XPS boards with W or MG surface, mechanical fasteners, base reinforcing layer with glass-fibre mesh, plaster primer, decorative plaster (render); the same sequence with primer and tile adhesive in place of render applies to tile-clad systems [S1]. The mechanical key provided by the W or MG face is what allows the base-coat to lock into the foam, with the glass-fibre mesh bridging thermal movement and the primer sealing the base-coat for the decorative finish.
Saint-Gobain Weber's technical guidance on correct rendering of XPS-R insulation boards mirrors the same logic: depending on the type of underlay render and reinforcing mortar, the surface is left as required for the post-applied topcoat, including a scratch-render option to add further key before the final coat [S7]. Edge profile also matters: square edges demand tighter cutting and joint control, while shiplap or rabbet edges reduce gaps and airflow paths; tongue-and-groove edge XPS such as DuPont Styrofoam is available in 3/4 in. to 2 in. thicknesses and 2 ft. or 4 ft. by 8 ft. board sizes with R-values from 3.8 to 10 per inch-based rating per ASTM C518 [S4].
Who this is for, and who it is not for
W or MG XPS is the right call for specifiers writing ETICS render systems, tile-on-XPS facades, plinth insulation with an applied finish, and any project where a cementitious compound contacts the board face; smooth XPS is the right call for foundation and basement work, slab-on-grade, inverted-roof build-ups, road and infrastructure construction, cavity-wall insulation, and expansion-joint fillers [S3]. The same decision logic applies in adjacent bonded-laminate selection, for example in slit sheet versus cold-drawn wire steel fibre choice where surface preparation drives bond performance in a comparable way.
It is not for projects that need a single universal board across the entire envelope: a project that uses XPS in both foundation walls and a rendered ETICS facade will legitimately need two different surface types on the same site, and a single-line spec that defaults to smooth will fail on the facade and a single-line spec that defaults to W or MG will over-spec and over-cost on the foundations. The PNP product range is structured for exactly this split, with smooth grades (XPS 300, XPS 500, XPS 700) for buried applications and W or MG grades (XPS 150/200/300 W or MG) for rendered and tiled systems [S3].
Standards, testing, and sourcing
All three PNP surface variants are declared under EN 13164:2012+A1:2015, with thermal conductivity tested to the harmonised product standard and surface-specific water-absorption classes WL(T)0.7 for smooth and W, WL(T)3 for MG [S3]. DuPont Styrofoam tongue-and-groove XPS, by contrast, is a U.S.-market Type IV product tested to ASTM C578, with thermal resistance of 5.0 ft²·h·°F/Btu per inch minimum by ASTM C518, compressive strength of 25 psi minimum by ASTM D1621, water absorption of 0.1% by volume maximum by ASTM C272, and water-vapour permeance of 1.5 perm maximum by ASTM E96 [S4]. Fire performance for that product is flame spread ≤ 25 and smoke developed ≤ 450 per UL 723 [S4].
Specifiers should write the purchase spec to include thickness, declared lambda value and test standard, compressive strength class, water-absorption requirement, dimensional-stability requirement, length/width/thickness tolerances, and explicit surface and edge type, because "same name, different quality" boards in the market are the usual source of disputes between supplier, contractor, and consultant [S5]. For broader bonded-system context where the substrate surface drives the adhesive choice, the industrial vs consumer adhesive gap is a useful parallel on how the substrate side of the joint is often under-specified relative to the adhesive.
Track the following two signals over the next procurement cycle: whether EN 13164 revision work shifts the WL(T)3 threshold (current ≤ 3.0 Vol.-% is the binding number for MG boards and is the value that lets render suppliers accept the surface in ETICS documentation), and whether more manufacturers publish explicit "do not render" language for smooth boards in line with PNP's current exclusion, which would make smooth-on-render a spec error visible at tender stage rather than at facade inspection [S1][S3].
The underlying component specifications are covered under xps board, surface roughness tester, and eps board.